Disinfection Byproducts Acrylamide. Food Safety Applications Notebook Processing Contaminants

Transcription

1 Disinfection Byproducts Acrylamide Food Safety Applications Notebook Processing Contaminants

2 Table of Contents Index of Analytes and Application Notes... 3 Introduction to Food Safety... 4 UltiMate 3000 UHPLC + Systems... 5 IC and RFIC Systems... 6 MS Instruments... 7 Chromeleon 7 Chromatography Data System Software... 8 Process Analytical Systems and Software... 9 Automated Sample Preparation Analysis of Processing Contaminants Fast Determination of Acrylamide in Food Samples Using Accelerated Solvent Extraction (ASE) Followed by Ion Chromatography with UV or MS Detection Extraction and Cleanup of Acrylamide in Complex Matrices Using Accelerated Solvent Extraction (ASE) Followed by Liquid Chromatography Tandem Mass Spectrometry (LC-MS/MS) Determination of Inorganic Oxyhalide Disinfection Byproduct Anions and Bromide in Drinking Water Using Ion Chromatography with the Addition of a Postcolumn Reagent for Trace Bromate Analysis Determination of Bromate in Bottled Mineral Water Using the CRD 300 Carbonate Removal Device Determination of Trace Concentrations of Chlorite, Bromate, and Chlorate in Bottled Natural Mineral Waters Determination of Trace Concentrations of Oxyhalides and Bromide in Municipal and Bottled Waters Using a Hydroxide-Selective Column with a Reagent-Free Ion Chromatography System Determination of Polycyclic Aromatic Hydrocarbons (PAHs) in Edible Oils by Donor-Acceptor Complex Chromatography (DACC)-HPLC with Fluorescence Detection Robust and Fast Analysis of Tobacco-Specific Nitrosamines by LC-MS/MS Determination of Hydroxymethylfurfural in Honey and Biomass Column Selection Guide... 78

3 Index of Analytes and Application Notes Analytes Acrylamide... 13, 17 Bromate... 30, 38 Bromide... 21, 46 Chlorate Chlorite Hydroxymethylfurfural Tobacco-specific nitrosamines Oxyhalide PAHs Application Note Index Application Notes by Number Application Note Application Note Application Note Application Note Application Note Application Note Application Note Application Note Application Note Index

4 Introduction to Food Safety Food contamination stories in the news media have raised awareness of the fact that we live with a global food supply chain, and food safety is increasingly becoming an important concern. All types of fruits, vegetables, seafood, and meat can be purchased year round independent of the local growing season. For example, in many countries, well-stocked grocery stores carry cantaloupes from Guatemala, cucumbers from Mexico, shrimp from Vietnam, and fish from China. With fruit, vegetables, seafood, and meat traveling thousands of miles to reach far-flung destinations, and with poor or no knowledge of the agricultural practices, the need for food testing is increasingly important. Thermo Fisher Scientific understands the demands of food safety related testing. Our separation and detection technologies, combined with experienced applications competence, and our best suited chemistries provide solutions for the analysis of inorganic ions, small drug molecules, pesticides to large components, such as polysaccharides. Your laboratory can now conduct reliable, accurate, and fast testing of food. This notebook contains a wide range of food safety related application notes that will help address your food safety issues. Thermo Scientific and Dionex Integrated Systems Dionex Products are now a part of the Thermo Scientific brand, creating exciting new possibilities for scientific analysis. Now, leading capabilities in liquid chromatography (LC), ion chromatography (IC), and sample preparation are together in one portfolio with those in mass spectrometry (MS). Combining Dionex s leadership in chromatography with Thermo Scientific s leadership position in mass spec, a new range of powerful and simplified workflow solutions now becomes possible. For more information on how the new lineup of Thermo Scientific products can expand your capabilities and provide the tools for new possibilities, choose one of our integrated solutions: Ion Chromatography and Mass Spectrometry Liquid Chromatography and Mass Spectrometry Sample Preparation and Mass Spectrometry 4 Introduction

5 UltiMate 3000 UHPLC + Systems Best-in-class HPLC systems for all your chromatography needs Thermo Scientific Dionex UltiMate 3000 UHPLC + Systems provide excellent chromatographic performance while maintaining easy, reliable operation. The basic and standard analytical systems offer ultra HPLC (UHPLC) compatibility across all modules, ensuring maximum performance for all users and all laboratories. Covering flow rates from 20 nl/min to 10 ml/min with an industry-leading range of pumping, sampling, and detection modules, UltiMate 3000 UHPLC + Systems provide solutions from nano to semipreparative, from conventional LC to UHPLC. Superior chromatographic performance UHPLC design philosophy throughout nano, standard analytical, and rapid separation liquid chromotography (RSLC) 620 bar (9,000 psi) and 100 Hz data rate set a new benchmark for basic and standard analytical systems RSLC systems go up to 1000 bar and data rates up to 200 Hz 2 Dual System for increased productivity solutions in routine analysis Fully UHPLC compatible advanced chromatographic techniques Thermo Scientific Dionex Viper and nanoviper the first truly universal, fingertight fitting system even at UHPLC pressures Thermo Fisher Scientific is the only HPLC company uniquely focused on making UHPLC technology available to all users, all laboratories, and for all analytes. Rapid Separation LC Systems: The extended flowpressure footprint of the RSLC system provides the performance for ultrafast high-resolution and conventional LC applications. RSLCnano Systems: The Rapid Separation nano LC System (RSLCnano) provides the power for highresolution and fast chromatography in nano, capillary, and micro LC. Standard LC Systems: Choose from a wide variety of standard LC systems for demanding LC applications at nano, capillary, micro, analytical, and semipreparative flow rates. Basic LC Systems: UltiMate 3000 Basic LC Systems are UHPLC compatible and provide reliable, highperformance solutions to fit your bench space and your budget. 5 Liquid Chromatography Systems

6 IC and RFIC Systems A complete range of ion chromatography solutions for all customer performance and price requirements For ion analysis, nothing compares to a Thermo Fisher Scientific ion chromatography system. Whether you have just a few samples or a heavy workload, whether your analytical task is simple or challenging, we have a solution to match your needs and budget. And with your IC purchase, you get more than just an instrument you get a complete solution based on modern technology and world-class support. Thermo Scientific Dionex ICS-5000: The world s first capillary IC system Dionex ICS-2100: Award-winning integrated Reagent-Free IC system Dionex ICS-1600: Standard integrated IC system Dionex ICS-1100: Basic integrated IC system Dionex ICS-900: Starter line IC system Ranging from the Dionex ICS-900 to the ICS-5000, these IC systems cover the entire range of IC needs and budgets and come with superior support and service worldwide. Dionex ICS-5000: Developed with flexibility, modularity, and ease-of-use in mind, the Dionex ICS-5000 combines the highest sensitivity with convenience Dionex ICS-2100: An integrated Reagent-Free IC (RFIC ) system for electrolytically generated isocratic and gradient separations with conductivity detection, now with electrolytic sample preparation. Dionex ICS-1600: The Dionex ICS-1600 combines high sensitivity with convenience. Now ready for eluent regeneration, with available dual-valve configuration for automated sample preparation. Dionex ICS-1100: With dual-piston pumping and electrolytic suppression. Now ready for eluent regeneration, with available dual-valve configuration for automated sample preparation. Dionex ICS-900: Can routinely analyze multiple anions and cations in min fully automated with Displacement Chemical Regeneration (DCR). 6 Ion Chromatography Systems

7 MS Instruments Single-point control and automation for improved easeof-use in LC/MS and IC/MS Thermo Fisher Scientific provides advanced integrated IC/MS and LC/MS solutions with superior ease-of-use and modest price and space requirements. UltiMate 3000 System Wellness technology and automatic MS calibration allow continuous operation with minimal maintenance. The Dionex ICS-5000 instrument and the family of RFIC systems automatically remove mobile phase ions for effort-free transition to MS detection. Thermo Scientific MSQ Plus mass spectrometer, the smallest and most sensitive single quadrupole on the market for LC and IC Self-cleaning ion source for lowmaintenance operation Thermo Scientific Dionex Chromeleon Chromatography Data System software for single-point method setup, instrument control, and data management Compatible with existing IC and LC methods The complete system includes the MSQ Plus mass spectrometer, PC datasystem, electrospray ionization (ESI) and atmospheric pressure chemical ionization (APCI) probe inlets, and vaccum system You no longer need two software packages to operate your LC/MS system. Chromeleon LC/MS software provides single-software method setup and instrument control; powerful UV, conductivity, and MS data analysis; and fully integrated reporting. MS Systems and Modules: MSQ Plus Mass Spectrometer; MSQ18LA nitrogen gas generator; Thermo Scientific Dionex AXP-MS digital auxiliary pump 7 MS Instruments

8 Chromeleon 7 Chromatography Data System Software The fastest way to get from samples to results Discover Chromeleon software version 7, the chromatography software that streamlines your path from samples to results. Get rich, intelligent functionality and outstanding usability at the same time with Chromeleon software version 7 the Simply Intelligent chromatography software. Enjoy a modern, intuitive user interface designed around the principle of operational simplicity Streamline laboratory processes and eliminate errors with eworkflows, which enable anyone to perform a complete analysis perfectly with just a few clicks Access your instruments, data, and eworkflows instantly in the Chromeleon Console Locate and collate results quickly and easily using powerful built-in database query features Interpret multiple chromatograms at a glance using MiniPlots Find everything you need to view, analyze, and report data in the Chromatography Studio Accelerate analyses and learn more from your data through dynamic, interactive displays Deliver customized reports using the built-in Excelcompatible speadsheet Chromeleon software version 7 is a forward-looking solution to your long-term chromatography data needs. It is developed using the most modern software tools and technologies, and innovative features will continue to be added for many years to come. The Cobra integration wizard uses an advanced mathematical algorithm to define peaks. This ensures that noise and shifting baselines are no longer a challenge in difficult chromatograms. When peaks are not fully resolved, the SmartPeaks integration assistant visually displays integration options. Once a treatment is selected, the appropriate parameters are automatically included in the processing method. Chromeleon software version 7 ensures data integrity and reliability with a suite of compliance tools. Compliance tools provide sophisticated user management, protected database stuctures, and a detailed interactive audit trail and versioning system. 8 Chromatography Data Systems

9 Process Analytical Systems and Software Improve your process by improving your process monitoring with a Thermo Scientific Dionex on-line IC or HPLC system Our process analytical systems provide timely results by moving liquid chromatography-based measurements on-line. Information from the Thermo Scientific Dionex Integral process analyzer can help reduce process variability, improve efficiency, and reduce downtime. These systems provide comprehensive, precise, accurate information faster than is possible with laboratory-based results. From the lab to the factory floor, your plant s performance will benefit from the information provided by on-line LC. Characterize your samples completely with multicomponent analysis Reduce sample collection time and resources with automated multipoint sampling Improve your process control with more timely results See more analytes with unique detection capabilities 25 years of experience providing on-line IC and HPLC capabilities to a wide range of industries The Thermo Scientific Integral Migration Path approach lets you choose the systems that best meets your needs The Integral Migration Path approach enables on-line IC/HPLC to generate timely, high-resolution information when monitoring a small-scale reactor in a process R&D lab, in a pilot plant, or improving current manufacturing plant processes. No matter what the application, the Integral process analyzer has the versatility to place a solution using on-line IC/HPLC, whenever and wherever it is needed. Integral: The Integral Migration Path approach: System solutions wherever you need them: lab, pilot plant, or manufacturing Chromeleon Process Analytical (PA) Software: Chromeleon PA software provides unique capabilities to support on-line IC or HPLC analysis 9 Process Analytical Systems and Software

10 Automated Sample Preparation Accelerated Solvent Extractors Two new solvent extraction systems with ph-hardened Dionium components We offer two solvent extraction systems. The Thermo Scientific Dionex ASE 150 Accelerated Solvent Extractor is an entry-level system with a single extraction cell, for laboratories with modest throughput. The Dionex ASE 350 system is a sequential extraction system capable of automated extraction of up to 24 samples. Both systems feature chemically inert Dionium components that allow the extraction of acid- or basepretreated samples. 10 Automated Sample Preparation

11 Solid-Phase Extraction Systems Faster, more reliable solid-phase extraction while using less solvent The Thermo Scientific Dionex AutoTrace 280 Solid- Phase Extraction (SPE) instrument unit can process six samples simultaneously with minimal intervention. The instrument uses powerful pumps and positive pressure with constant flow-rate technology. Current analytical methods that require SPE sample preparation include gas chromatography (GC), GC-MS, LC, and LC-MS, IC and IC-MS. The Dionex AutoTrace 280 instrument is approved or adapted for U.S. EPA clean water methods and safe drinking water methods (600 and 500 series) and can extract the following analytes: PCBs (polychlorinated biphenyls) OPPs (organophosphorus pesticides), OCPs (organochlorine pesticides), and chlorinated herbicides BNAs (base, neutral, acid semivolatiles) Dioxins and furans PAHs (polyaromatic hydrocarbons) Oil and grease or hexane extractable material With SPE, large volumes of liquid sample are passed through the system and the compounds of interest are trapped on SPE adsorbents (cartridge or disk format), then eluted with strong solvents to generate an extract ready for analysis. Automated SPE saves time, solvent, and labor for analytical laboratories. Dionex AutoTrace Systems: The new Dionex AutoTrace 280 system provides fast and reliable automated solid phase extraction for organic pollutants from liquid samples Dionex AutoTrace Accessories: High-quality parts and accessories are available for Dionex AutoTrace 280 instruments 11 Automated Sample Preparation

12 Analysis of Processing Contaminants

13 Application Note 409 Fast Determination of Acrylamide in Food Samples Using Accelerated Solvent Extraction (ASE ) Followed by Ion Chromatography with UV or MS Detection Introduction Acrylamide, a known genotoxic compound, was recently detected in carbohydrate-rich fried or baked food samples by a Swedish research group, Tareke, et al. 1 The content of acrylamide was as high as several mg/kg for typical samples such as hash browns and french fries. Published methods for acrylamide include U.S. EPA Method 8032A that uses liquid extraction and GC-ECD for determinations in water 2 and a method by the German Health Agency (BGVV) that uses HPLC with UV detection for migration analysis of acrylamide from food packing materials. 3 The method presented here consists of a fast, automated extraction method using accelerated solvent extraction (ASE). 4 Samples were extracted in 20 min using pure water, water with 10 mm formic acid, or acetonitrile. The extracts were directly analyzed by ion chromatography (IC) using a 4-mm ion-exclusion column and both UV and MS detection. With this column, acrylamide is retained longer than on conventional reversed-phase columns, allowing separation from the many coextractable compounds present in food samples. Results were obtained for acrylamide in french fries, potato chips, and crisp bread. The benefits of this method are simplicity, speed of analysis, and a degree of automation that allows the analysis of large numbers of samples with minimal labor. Conditions Extraction Conditions Solvent: Water, 10 mm formic acid, or acetonitrile Temperature: 80 C Pressure: 10 MPa Heatup Time: 5 min Static Time: 4 min Number of Static Cycles: 3 Flush Volume: 60% Purge Time (N 2 ): 120 s Chromatographic Conditions Column: IonPac ICE-AS1, mm, 7.5 µm, SP6003 Eluent: 3.0 mm formic acid in acetonitrile/water 30/70 (v/v) Flow Rate: 0.15 ml/min Inj. Volume: 25 µl UV Detection: 202 nm MS Detection: ESI+: 3.0 kv, cone 50 V, probe temp. 300 C, scan m/z, SIM 72 m/z 13 Fast Determination of Acrylamide in Food Samples Using Accelerated Solvent Extraction (ASE) Followed by Ion Chromatography with UV or MS Detection

14 Experimental Extraction Samples of 5 g were extracted using an accelerated solvent extraction system, (ASE 100 or ASE 200, Dionex, Sunnyvale, CA, USA) with 34-mL cells for the ASE 100, and 33-mL cells for the ASE 200. Chromatography Chromatographic analyses were performed on an DX-600 ion chromatograph, (Dionex, Sunnyvale, CA, USA) that included a GS50 gradient pump, PDA-100 photodiode array detector set at 202 nm, an MSQ single quadrupole mass spectrometer, and an AS50 Autosampler. A mm i.d. IonPac ICE-AS1 (Dionex, Sunnyvale, CA, USA) analytical column (7.5-µm cross-linked polystyrene divinylbenzene functionalized with sulfonate functional groups) was used to separate acrylamide from the matrix compounds. All measurements were made at 30 C and all samples were filtered through 0.45-µm filters. A 25-µL sample loop was used for all the determinations. Data collection and the operation of all components in the system was controlled by Dionex Chromeleon 6.40 chromatography software. Reagents and Standards All reagents were analytical-grade. Formic acid was Suprapur (Merck, Darmstadt, Germany), and acetonitrile was HPLC reagent-grade (Novachimica, Milano, Italy). Ultrapure water with conductivity <0.1 ΩS (DI water) was obtained from a Milli-Q system (Millipore, Bedford, MA, USA). Working standard solutions of acrylamide were prepared by serial dilution of a 1000-mg/L stock standard solution. Samples French fries, potato chips, tortilla chips, wheat snacks with bacon flavor, and crisp bread were obtained from a local food store. Representative samples (5 g) were loaded into 34-mL extraction cells onto a glass fiber filter. Samples like wheat snacks with bacon flavor or bread samples, which have a tendency to dissolve or swell, were loaded into Soxhlet thimbles that were then placed in the extraction cells. Any void volumes were filled with glass beads (1-mm i.d.) to reduce the volume of the extraction solvent. Results and Discussion ASE Extraction Pure water, water with 10 mm formic acid, and acetonitrile were tested as the extraction solvent. Pure water extracts showed lower recoveries than the formic acid, but the formic acid extracts had lower stability. Extracts produced using acetonitrile were cleaner, as less material was coextracted from the sample matrix. The extraction temperature of 80 C was chosen, because acrylamide starts to decompose at temperatures above 83 C. With three extraction cycles of 4-min durations, a spiked french fries sample had a yield of 95% in the first extract and an additional 8% in the second extraction of the same sample using 10 mm formic acid. Cleanup The extract volume was determined using a volumetric flask. Afterward, the extracts were filtered using a 0.22-µm nylon filter. Further cleanup using solid phase extraction or liquid extraction did not exhibit any significant improvements for the subsequent chromatographic analyses. Analysis of Acrylamide Using IC/UV The separation of acrylamide was performed using an IC system with a UV detector. Formic acid was chosen as the eluent instead of sulphuric acid because it is more compatible with MS detection. The amount of acetonitrile was optimized to 30% v/v to reduce the total run time and avoid interferences with matrix components. 14 Fast Determination of Acrylamide in Food Samples Using Accelerated Solvent Extraction (ASE) Followed by Ion Chromatography with UV or MS Detection

15 mau 1 mau B A Figure 1. Comparison of french fry samples. Chromatographic conditions are listed under Conditions. UV detection. (a) neat; (b) spiked with 40 µg of acrylamide. Peak: 1. Acrylamide. Figure 3. Chromatogram of a crisp bread sample with low acrylamide content. Chromatographic conditions are listed under Conditions. UV detection. Peaks: arrow indicates acrylamide retention time Intensity m/z Figure 2. Mass spectrum of acrylamide in the range m/z. Mass spectrometric conditions are listed under Conditions. Figure 4. Chromatogram of a crisp bread sample with low acrylamide content. Chromatographic conditions are listed under Conditions. MS detection SIM mode. Peak: 1. Acrylamide 0.08 mg/kg. Analysis of Acrylamide Using IC/MS A single-stage quadrupole mass spectrometric detector (Thermo Election MSQ Dionex, Sunnyvale, CA) was installed in series with the UV detector. The MS was operated in the positive electrospray (ESI+) ionization mode. Figure 2 shows a mass spectrum of acrylamide at a cone voltage of 50 V. The protonated molecular ion [M+H] + of acrylamide is detected at a mass-to-charge ratio (m/z) 72. In addition, a fragment ion is observed at m/z 55. This lower-intensity fragment ion can be used for confirmation while a more sensitive detection is achieved at m/z 72. Calibrations were performed using external standards in the range mg/l. The corresponding calibration plots for both UV and MS detection show good linearity in the range mg/l (r 2 = 0.996). Conclusion The ASE method provides a fast and efficient extraction of acrylamide from various food samples. The extracted samples were analyzed directly using IC with UV or MS detection. Although UV detection is sufficient for most of the analyzed samples, MS detection offers a higher specificity and sensitivity, as shown in Figures 3 6. Results are summarized in Table 1. The required limits of determination of 50 µg/kg acrylamide in food can be achieved with this method. This method is robust, selective, and relatively easy to perform. 15 Fast Determination of Acrylamide in Food Samples Using Accelerated Solvent Extraction (ASE) Followed by Ion Chromatography with UV or MS Detection

16 100 mau 10 Figure 5. Chromatogram of a potato chips sample with high acrylamide content. Chromatographic conditions are listed under Conditions. UV detection. Peak: 1. Acrylamide 1.56 mg/kg Intensity Figure 6. Chromatogram of a potato chips sample with high acrylamide content. Chromatographic conditions are listed under Conditions. MS detection SIM mode. Peak: 1. Acrylamide 1.06 mg/kg Sample Table 1. Acrylamide Contents of Food Samples Acrylamide UV* mg/kg Acrylamide MS** mg/kg Chips bacon flavored n.d. <0.05 Crisp bread n.d Potato chips Tortilla chips n.d. n.d. French fries French fries spiked 0.70 a 0.69 b * 202 nm; ** SIM 72 m/z; a recovery 96.4%; b recovery 95.1% References 1. Tareke, E.; Ryberg, P.; Karlsson, P.; Eriksson, S.; Tornqvist, M., J. Agric. Food. Chem. 2002, 50, U.S. EPA Method 8032A; SW-846 Rev 1; U.S. Environmental Protection Agency, Office of Solid Waste, December Acrylamide in food serious problem or exaggerated risk? Results of BgVV Information Seminar, 29 Aug 2002; foods serious_problem_or_ exaggerated_risk.pdf 4. Richter, B. E.; Jones, B. A.; Ezzell, J. L.; Porter, N. L.; Avdalovic, N.; Pohl, C. Anal. Chem. 1996, 68, Fast Determination of Acrylamide in Food Samples Using Accelerated Solvent Extraction (ASE) Followed by Ion Chromatography with UV or MS Detection

17 Application Note 358 Extraction and Cleanup of Acrylamide in Complex Matrices Using Accelerated Solvent Extraction (ASE ) Followed by Liquid Chromatography Tandem Mass Spectrometry (LC-MS/MS) Introduction Acrylamide is formed during the cooking process of certain plant-based foods which are rich in carbohydrates and low in protein. Specifically, it forms when asparagine reacts with sugars such as glucose at high temperatures. Acrylamide was detected in fried foods by the Swedish National Food Authority in Since then, many food laboratories have successfully performed determinations for this compound on a variety of different food matrices. Acrylamide is a known carcinogen in animals. ASE is an excellent technique for extraction of acrylamide from various fried food products; until recently, however, extraction of this compound from matrices such as coffee and chocolate has proven difficult. Traditional extraction techniques are time consuming, and may cause bottlenecks in sample preparation. This Application Note describes a new ASE method that combines the extraction of low-levels of acrylamide from coffee and chocolate with an in-cell, solid-phase cleanup step. Equipment Dionex ASE 200 with 33-mL stainless steel extraction cells (P/N ) Dionex cellulose filters (P/N ) Dionex collection vials 60 ml (P/N ) Dionex SE 500 Solvent Evaporator (P/N , 120 v) (P/N , 240 v) Standard laboratory tissue homogenizer Standard laboratory centrifuge (rated to 10,000 rpm or greater) Centrifuge tubes (40 50 ml) Chemicals and Reagents Acrylamide, purity 99% (Sigma-Aldrich) d3-acrylamide (2,3,3-d3-2-propenamide) (Cambridge, Isotope Laboratories USA) Florisil, mesh (Sigma-Aldrich) Potassium hexacyanoferrate (II) trihydrate (Carrez I) (Sigma-Aldrich) Zinc sulfate heptahydrate (Carrez II) (Sigma-Aldrich) Dionex ASE Prep DE (P/N ) Termamyl 120 L (Type L thermostable amyloglucosidase enzyme) (Novozymes, Denmark) Ethyl acetate (Fisher Scientific, HPLC Grade) Dichloromethane (Fisher Scientific, HPLC Grade) Methanol (Fisher Scientific, HPLC Grade) 17 Extraction and Cleanup of Acrylamide in Complex Matrices Using Accelerated Solvent (ASE) Followed by Liquid Chromatography Tandem Mass Sepctrometry (LC-MS/MS)

18 Food Samples The coffee and chocolate samples were purchased from a local grocery store and stored at room temperature. Reagent Solutions 0.68 M potassium hexacyanoferrate (II) trihydrate (Carrez I) solution Dissolve g of K 4 Fe(CN)6 3H 2 O in 100 ml of water. 2 M zinc sulfate heptahydrate (Carrez II) solution Dissolve g of ZnSO 4 7H 2 O in 100 ml of water. Standard Solutions Prepare aqueous stock solutions of acrylamide and d3-acrylamide at concentrations of 50 and 5 µg/ml, respectively. To make 50 µg/ml acrylamide, add 5 mg to 100 ml water. To make 5 µg/ml d3-acrylamide, add 0.5 mg to 100 ml water. Dilute the acrylamide solutions in water to obtain the following matrix-equivalent levels: 0, 10, 50, 200, 500, and 2500 µg/kg. The matrix-equivalent concentration of d3-acrylamide should be 250 µg/kg. Sample Preparation Hydrolysis Weigh 2.0 g of sample into a centrifuge tube and add 10 ml of water, (heated to 60 C) then add 50 µl of Termamyl. Place the tube in a water bath at 90 C for 45 min. Homogenize the mixture for 1 min. To precipitate the proteins, add 1 ml of the 0.68 M potassium hexacyanoferrate (II) trihydrate solution and 1 ml of the 2 M zinc sulfate heptahydrate solution to the centrifuge tube, swirling constantly for 1 2 min. Add 5 ml dichloromethane and swirl for an additional minute. Centrifuge for 15 min at 10,000 rpm. Preparing the ASE Cell Prepare the 33-mL extraction cell by successively inserting: (1) a cellulose filter, (2) 6 g Florisil deactivated with 3% deionized water, (3) a second cellulose filter, and (4) at least 8 g ASE Prep DE so that there is approximately 0.5 cm of empty space at the top of the cell. Transfer 6 ml of the supernatant from the centrifuge tube and drip onto the ASE Prep DE layer in the prepared extraction cell. Fill the extraction cell to the top with additional ASE Prep DE and cover with a third cellulose filter. Carefully place the extraction cell cap on the cell and tighten by hand. ASE Conditions Solvent: Ethyl acetate (100%) Temperature: Ambient* Pressure: 2000 psi Static Time: 3 min Static Cycles: 3 Flush: 100% Purge: 60 s * Although the authors of this Application Note used ambient temperature for the extraction, this is not a typical extraction temperature for ASE. Normal extraction temperatures range from C. If recoveries are lower than expected using ambient temperature, increasing temperature may improve results. Extraction Place the prepared extraction cells onto the ASE carousel. Enter the ASE conditions into the Method Editor screen, and save this method with the desired number. Begin the extraction by pushing Start. The method can also be set up as a Schedule in the Schedule Editor screen. Running the ASE system under Schedule control enables the system to track any errors that may occur throughout an extraction run. This is especially helpful if the system is set up to run unattended overnight. Any problems are logged in the Error Log for the user to view the next morning. When the extraction is complete, evaporate the extracts under vacuum (40 C, 200 mbar) until only a few droplets remain. Evaporate the residual ethyl acetate under a gentle stream of nitrogen, or transfer the vials to the Dionex SE 500 Solvent Evaporation System (P/N ~120v; ~240v) and evaporate to dryness using standard conditions. Redissolve the extract with 500 µl water. If necessary, the extract can be filtered through a 0.2-µm cellulose filter. Decant approximately 180 µl of the extract and mix with 90 µl of methanol prior to analysis by LC-MS/MS. 18 Extraction and Cleanup of Acrylamide in Complex Matrices Using Accelerated Solvent (ASE) Followed by Liquid Chromatography Tandem Mass Sepctrometry (LC-MS/MS)

19 Results and Discussion The ASE method automates the extraction and cleanup steps of extraction of acrylamide from cocoa and coffee. Compared to the manual method, ASE greatly reduces the time and the amount of sample handling required (Table 1). The addition of Florisil to the extraction cell eliminates the need for an additional cleanup step of the extract. Figure 1 shows coffee extracted with various amounts of Florisil, followed by filtration through SPE cartridges. Method optimization determined that 6 g of Florisil were sufficient to obtain a clear extract, however some samples may require additional filtering before analysis. Figure 1. Residual coextractables from coffee extracts, trapped on an Isolute Multimode SPE cartridge. The four extraction cells contained from 0 to 6 g of Florisil. From left to right: (1) no Florisil, (2) 2 g Florisil, (3) 4 g Florisil, and (4) 6 g Florisil. Table 1. Comparison of the Sample Preparation Steps for the Manual Method and ASE Method for Extraction of Acrylamide from Coffee and Chocolate Manual Method 1. Weigh 2 g of sample, add 100 µl of d3-acrylamide solution (5 µg/ml) and 10 ml of water into a centrifuge tube. Homogenize for 1 min. 2. Add 1 ml Carrez I, swirl, add 1 ml Carrez II, swirl. Add 5 ml dichloromethane, shake vigorously for 1 min. ASE Method 1. Weigh 2 g of sample, add 100 µl of d3-acrylamide solution (5 µg/ml) and 10 ml of water into a centrifuge tube. Homogenize for 1 min. 2. Add 1 ml Carrez I, swirl, add 1 ml Carrez II, swirl. Add 5 ml dichloromethane, shake vigorously for 1 min. 3. Centrifuge at 3 5 C, 10,000 rpm for 15 min. 3. Centrifuge at 3 5 C, 10,000 rpm for 15 min. 4. Transfer 6 ml of the supernatant into a centrifuge tube containing 1.8 g of NaCl, swirl to dissolve. 4. Prepare ASE cell by successively adding 1 cellulose filter, 6 g of Florisil (deactivated with 3% water), a second cellulose filter and 8 g ASE Prep DE. 5. Add 13 ml ethyl acetate and shake vigorously (1 min). 5. Transfer 6 ml of the extract on the ASE Prep DE, fill the rest of the column with ASE Prep DE, add a third cellulose filter and close the cell. 6. Centrifuge at 3 5 C, g for 15 min. 6. Perform the ASE extraction step. 7. Transfer the organic phase into an amber vial containing 2 ml water. Shake vigorously for 1 min. 7. Evaporate the organic fraction under vacuum (40 C, 200 mbar) to about 500 µl and finish the evaporation with a gentle stream of N Evaporate the organic phase with N 2 at 40 C. 8. Redissolve the extract in 500 µl water. Repeat the ethyl acetate extraction (2 ), steps 5-7. Add 90 µl of methanol to 180 µl of extract. Proceed with LC-MS/ MS analysis (60 µl injected) 9. Condition the SPE cartridge with 3 ml methanol, then twice with 3 ml distilled water. 10. Load the aqueous extract onto the cartridge, elute and rinse with 1 ml water. collect both fractions. 11. Reduce the extract volume to approximately 500 µl (N 2, 40 C). 12. Add 90 µl of methanol to 180 µl of extract. Proceed with LC-MS/MS analysis (60 µl injected). 19 Extraction and Cleanup of Acrylamide in Complex Matrices Using Accelerated Solvent (ASE) Followed by Liquid Chromatography Tandem Mass Sepctrometry (LC-MS/MS)

20 Table 2 compares the results of manual extraction to ASE extraction of blank samples spiked with acrylamide standard. Table 2. Comparison of Manual Extraction versus ASE for Quantification of Acrylamide Spiked Samples in Soluble Chocolate Powder (n = 6) Manual Extraction ASE Extraction Spiking Levels Recovery % %RSD Recovery % %RSD 12.7 µg/kg µg/kg µg/kg Table 3 shows the results of using ASE for the extraction of acrylamide from various difficult matrices. Table 3. ASE of Roast Ground Coffee, Soluble Coffee, Coffee Surrogate, and Cocoa Acrylamide Level (µg/kg) Spiked at 150 µg/kg Materials Incurred a Expected b Measured a CV% R&G coffee Soluble coffee powder Coffee surrogate Cocoa powder a Mean of two independent determinations. b Mean incurred level + spike level. Conclusions ASE has consistently proven to be an excellent alternative to the traditional labor-intensive extraction methods used for determination of acrylamide in food. ASE allows extraction and cleanup to be performed simultaneously, eliminating the need for a post-extraction cleanup step. Automation of ASE allows for unattended extraction, and can be set up to run over night to provide the user with filtered extracts that are ready for analysis in the morning. The advantages of speed and decreased sample handling as compared to manual extraction techniques are clear. Acrylamide recoveries obtained with the ASE method were comparable to traditional methods, proving that ASE is an effective tool for the extraction of polar compounds from complex matrices. References 1. Wenzl, T., de la Calle, B. Anklam, E.; Analytical Methods for the Determination of Acrylamide in Food Products: A Review. Food Addit. Contam. 2003, 20, Delatour, T., Perisset, A., Goldmann, T., Riediker, S., Stadler, R. H.; Improved Sample Preparation to Determine Acrylamide in Difficult Matrixes Such as Chocolate Powder, Cocoa, and Coffee by Liquid Chromatography Tandem Mass Spectroscopy. J. Agric. Food Chem. 2004, 52, Cavalli, S., Maurer R., Hoefler F.; Fast Determination of Acrylamide in Food Samples Using Accelerated Solvent Extraction Followed by Ion Chromatography with UV or MS Detection. LC-GC International 2003, 16, Applications Book, Hoefler F., Maurer R., Cavalli S.; Schnelle Analyse von Acrylamid in Lebensmitteln mit ASE und LC/ MS, GIT Labor-Fachzeitschrift, 2002, 9, Suppliers Fisher Scientific, 2000 Park Lane, Pittsburgh, PA USA, Tel: , Sigma-Aldrich Chemical Company, 3050 Spruce St., St. Louis, MO USA, Tel: , Novozymes A/S, Krogshoejvej 36, 2880 Bagsvaerd, Denmark, Tel.: Fax: , ACKNOWLEDGMENT Thanks to the authors at the Nestle Research Center Ltd. and the Nestle Product Technology Center in Switzerland for allowing us to use their data for this application note. 20 Extraction and Cleanup of Acrylamide in Complex Matrices Using Accelerated Solvent (ASE) Followed by Liquid Chromatography Tandem Mass Sepctrometry (LC-MS/MS)

21 Application Note 136 Determination of Inorganic Oxyhalide Disinfection Byproduct Anions and Bromide in Drinking Water Using Ion Chromatography with the Addition of a Postcolumn Reagent for Trace Bromate Analysis Introduction The chlorination of drinking water can produce trihalo methanes and other suspected carcinogenic disinfection byproducts (DBPs) that endanger human health. 1 Unfortunately, common alternatives to chlorination also can produce harmful DBPs. The use of chlorine dioxide for the disinfection of drinking water generates the inorganic oxyhalide DBPs chlorite and chlorate, and the presence of chlorate has been reported in waters treated with hypochlorite. 2 Ozonation, an increasingly prevalent and effective disinfection technique, produces bromate as a DBP anion if the source water contains naturally occurring bromide. 3 Bromate has been judged by both the World Health Organization (WHO) and the U.S. Environmental Protection Agency (EPA) as a potential carcinogen, even at very low µg/l levels. The U.S. EPA has estimated a potential cancer risk equivalent to 1 in 10 4 for a lifetime exposure to drinking water containing bromate at 5 µg/l. 4 The U.S. EPA has recently issued new rules that require public water supplies to control previously unregulated microbes (e.g., cryptosporidium and giardia) and cancer-causing DBPs in finished drinking water. The Stage 1 D/DBP Rule specifies a Maximum Contaminant Level (MCL) for bromate of 10 µg/l and an MCL for chlorite of 1000 µg/l. 5 The EPA intends to convene Stage 2 of the D/DBP Rule in the near future, while both Germany and Japan are considering regulatory limits for inorganic DBPs. 6 The recent efforts by global regulatory agencies to monitor levels and establish regulatory limits has generated considerable interest in the development of improved analytical methods for the determination of trace level inorganic oxyhalide DBPs. The determination of bromate and other inorganic DBPs traditionally has been accomplished by ion chromatography (IC) using an IonPac AS9-SC anion-exchange column with a carbonate/ bicarbonate eluent and suppressed conductivity detection, as described in U.S. EPA Method (B). 7 EPA Method was published as an update to Method in Method specifies the use of an IonPac AS9-HC column and suppressed conductivity detection for the determination of bromate, bromide, chlorite, and chlorate at low µg/l levels by direct injection. 8 The detection limit for bromate determined by IC with suppressed conductivity detection can be further reduced to 1 µg/l by using preconcentration after appropriate sample cleanup. 1 Postcolumn derivatization can also be used to improve detection limits when using IC for inorganic DBP analysis. The use of IC with dual postcolumn addition of hydrochloric acid and then chlorpromazine can achieve a method detection limit (MDL) for bromate of 0.49 µg/l. 9 Iodate, chlorite, and bromate have been detected by using a postcolumn reaction with excess bromide under acidic conditions. The tribromide ion formed can be detected spectrophotometrically at 267 nm, allowing an MDL of less than 0.5 µg/l for bromate 21 Determination of Inorganic Oxyhalide Disinfection Byproduct Anions and Bromide in Drinking Water Using Ion Chromatography with the Addition of a Postcolumn Reagent for Trace Bromate Analysis

22 with a large-volume injection. 10 Sub-µg/L MDLs for bromate have also been reported by workers using other postcolumn reagents, such as fuchsin or excess iodide under acidic conditions. 11,12 In addition to postcolumn reaction (PCR) methods, electrospray tandem mass spectrometry (MS-MS) and inductively coupled plasma mass spectrometry (ICP-MS) have been used as specific detection techniques for the ion chromatographic analysis of bromate. The use of electrospray MS-MS detection can achieve an MDL for bromate of approximately 0.1 µg/l; the use of ICP-MS detection has been reported to permit an MDL for bromate of 0.8 µg/l. 13,14 This Application Note describes an improved IC method to quantify low levels of oxyhalide DBP anions and bromide in reagent water, bottled water, and finished drinking water. The method uses an IonPac AS9-HC column and suppressed conductivity detection, followed by postcolumn addition of o-dianisidine (ODA) to enhance visible absorbance detection of the bromate ion. This method allows quantification of all the key oxyhalide anions and bromide at low µg/l levels by using conductivity detection, and the postcolumn addition of ODA followed by visible detection allows quantification of bromate down to 0.5 µg/l. This method requires only a single postcolumn reagent delivered pneumatically with conventional postcolumn instrumentation. 2 The approach described in this Application Note is technically equivalent to that described in U.S. EPA Method titled Determination of Inorganic Oxyhalide Disinfection By- Products in Drinking Water Using Ion Chromatography with the Addition of a Postcolumn Reagent for Trace Bromate Analysis. 15 Equipment Dionex DX-500 ion chromatographic system consisting of: GP50 Gradient Pump with vacuum degas option ED40 Conductivity Detector with DS3 Detector Cell AD20 UV/Vis Absorbance Detector with 10-mm cell AS50 Autosampler PC10 Pneumatic Postcolumn Delivery Module (P/N 50601) PCH-2 Postcolumn Reaction Heater (P/N 39348) Knitted Reaction Coil, 500 µl, potted (for PCH-2) (P/N 39349) Two 4-L plastic bottle assemblies (for external water mode suppression) PeakNet 5.1 Chromatography Workstation Reagents and Standards Deionized water, Type I reagent grade, 18 MΩ-cm resistivity or better 0.5 M Carbonate Anion Eluent Concentrate (Dionex P/N 37162) o-dianisidine, dihydrochloride salt (ODA; Sigma D-3252) Iron (II) sulfate heptahydrate (Fe 2 SO 4 7H 2 O; Aldrich 21,542-2) Ethylenediamine (EDA; Sigma E-1521) Nitric acid, (70%; J.T. Baker INSTRA-ANALYZED ) Methanol (spectrophotometric grade; Sigma M-3641) Potassium bromide (KBr; J.T. Baker 2998) Sodium bromide (NaBr; Aldrich 31,050-6) Sodium bromate (NaBrO 3 ; EM SX ) Sodium chlorate (NaClO 3 ; Fluka 71370) Sodium chlorite (NaClO 2 ; Fluka 71388, ~80% pure) Bromate standard, 1000 mg/l, NaBrO 3 in H 2 O (SPEX CertiPrep AS-BRO39-2Y) Bromide standard, 1000 mg/l, NaBr in H 2 O (SPEX CertiPrep AS-BR9-2Y) Chlorate standard, 1000 mg/l, NaClO 3 in H 2 O (SPEX CertiPrep AS-CLO39-2Y) Chlorite standard, 1000 mg/l, NaClO 2 in H 2 O (SPEX CertiPrep AS-CLO29-2Y) Conditions Columns: Dionex AG9-HC, 50 4 mm ID guard column (P/N 51791) Dionex AS9-HC, mm ID analytical column (P/N 51786) Eluent: 9.0 mm Sodium carbonate (Na 2 CO 3 ) Flow Rate: 1.3 ml/min Temperature: Ambient Sample Volume: 225 µl Detection: Suppressed conductivity: ASRS -ULTRA (P/N 53946), AutoSuppression external water mode, 100 ma current, DS3 Cell (P/N 44130), 35 C, 1.7%/ C. Background Conductance: ~24 µs System Backpressure: ~2300 psi Run Time: 25 min 22 Determination of Inorganic Oxyhalide Disinfection Byproduct Anions and Bromide in Drinking Water Using Ion Chromatography with the Addition of a Postcolumn Reagent for Trace Bromate Analysis

23 PCR Detection: Postcolumn Reagent Flow: 0.7 ml/min Postcolumn Heater Temp.: 60 C Absorbance at 450 nm (tungsten lamp) Preparation of Solutions and Reagents Reagent Water Distilled or deionized water, 18 MΩ-cm or better, free of the anions of interest and filtered through a 0.2-µm filter. Eluent Solution (9 mm Sodium Carbonate) Dilute 18 ml of 0.5 M sodium carbonate concentrate to 1 L with deionized water. Unless the in-line degas option is being used, sparge eluent prior to use with helium or sonicate under vacuum for 10 min. Postcolumn Reagent Add 40 ml of 70% nitric acid to about 300 ml reagent water in a 500-mL volumetric flask. Add 2.5 g KBr and stir to dissolve. Dissolve 250 mg of o-dianisidine 2 HCl in 100 ml methanol and add to the nitric acid/kbr solution. Bring to volume with reagent water. Prepare in advance, set aside overnight until the slight champagne color fades, and filter through a 0.45-µm filter. Discard any PCR reagent that is not colorless or nearly colorless after sitting overnight. The reagent is stable for one month when stored at room temperature. Stock Standard Solutions Purchase certified solutions or prepare stock standard solutions by dissolving the corresponding mass of the salt for each of the anions of interest (see Table 1) in reagent water and dilute to 100 ml. Table 1 Masses of Compounds Used to Prepare 100 ml of 1000 mg/l Anion Standards Anion Compound Mass (g) - BrO 3 Sodium bromate (NaBrO 3 ) Br - Sodium bromide (NaBr) ClO 3 Sodium chlorate (NaClO 3 ) ClO 2 Sodium chlorite (NaClO 2 ) * *Because sodium chlorite is usually available only as an 80% technical grade salt, the 80% purity is accounted for in the g mass cited above. If an alternate purity is used, make an appropriate adjustment in the mass of salt used after determining the exact percentage of NaClO 2, which can be done using an iodometric titration procedure. 16 * Prepare a mixed anion calibration stock standard at 20 mg/l by combining 2 ml of each of the bromide, chlorite, and chlorate stock standards in a 100-mL volumetric flask. Mix and bring to volume with reagent water. These standards are stable for at least 1 month when stored at < 6 ºC. Because bromate decomposes in the presence of chlorite, prepare a bromate-only calibration stock standard at 5 mg/l by adding 0.5 ml of the bromate stock standard to a 100-mL volumetric flask and bringing to volume with reagent water. This standard is stable for 2 weeks when stored at < 6 C. Working Standard Solutions Use reagent water to prepare appropriate dilutions of the calibration stock standards as needed. Ethylenediamine (EDA) Preservative Solution Dilute 2.8 ml of ethylenediamine (99%) to 25 ml with reagent water. Prepare fresh monthly. Ferrous Iron Solution [1000 mg/l Fe (II)] Add 6 µl concentrated nitric acid to about 15 ml reagent water in a 25 ml volumetric flask. Add g ferrous sulfate heptahydrate (FeSO 4 7H 2 O), dissolve, and bring to volume with reagent water (final ph ~ 2). Prepare fresh every 2 days. Sulfuric Acid Solution (0.5 N) Dilute 1.4 ml of concentrated sulfuric acid to 100 ml with reagent water. Sample Preparation When taking a sample from a treatment plant that uses chlorine dioxide or ozone, the sample must be sparged immediately with an inert gas (e.g., nitrogen, argon, or helium) for 5 min. Add 1.00 ml of EDA Preservative Solution per 1.0 L of sample to prevent conversion of residual hypochlorite or hypobromite to chlorate or bromate. This also prevents metal-catalyzed conversion of chlorite to chlorate. The samples preserved in this manner are stable for at least 14 days when stored in amber glass bottles at 4 C. 17 After appropriate preservation, most samples can be filtered through a 0.45-µm filter and directly injected onto the ion chromatograph. However, each sample that contains excess chlorite must be treated to remove chlorite and then reanalyzed for bromate, because elevated levels of chlorite can interfere with the quantification of bromate by PCR. 23 Determination of Inorganic Oxyhalide Disinfection Byproduct Anions and Bromide in Drinking Water Using Ion Chromatography with the Addition of a Postcolumn Reagent for Trace Bromate Analysis

24 The treatment procedure to remove chlorite requires two portions of sample. Place two 10-mL aliquots of the sample into separate 20-mL beakers. Fortify one aliquot with bromate at a level approximating the native concentration of bromate in the untreated sample. This laboratory fortified matrix (LFM) will indicate correct performance of the chlorite removal step. Acidify both aliquots with 33 µl of sulfuric acid reagent and confirm the final ph (5 6) with ph test strips. Add 40 µl of ferrous iron solution, mix, and allow to react for 10 min. Filter the treated samples through a 0.45-µm nylon filter to remove precipitated ferric hydroxide, and then pass the solution through a hydronium form cation-exchange cartridge (Dionex OnGuard -H, P/N 39596) to remove excess soluble iron. The treated samples must be analyzed within 30 h. System Preparation and Set-up Configure the IC with the PCR system as depicted in Figure 1. Determine the PCR flow rate by collecting the combined effluent from the IC pump and the PCR module in a 10-mL graduated cylinder for 1 min. The PCR flow rate is the difference between the total flow rate and that of the IC pump. Adjust the air pressure of the postcolumn delivery module (PC10) and remeasure the flow rate until the correct flow rate of 0.7 ml/min is established. Confirm this flow rate on a weekly basis or whenever detector response for a calibration check standard deviates beyond quality control acceptance criteria. To determine target anions at trace concentrations, it is essential to have low baseline noise. Minimize baseline noise by taking the following steps during system set-up. Install the ASRS-ULTRA in the external water mode rather than the recycle mode. Prior to sample analysis, determine a system blank by analyzing 225 µl of deionized water using the method described above. An equilibrated system has a background conductance of ~ 24 µs, peak-to-peak noise of ~ 5 ns per minute, and no peaks eluting at the same retention time as the anions of interest. Results and Discussion Figure 2 shows the chromatograms of a mixed anion standard containing 10 µg/l bromate and 15 µg/l each of chlorite, bromide, and chlorate obtained by using dual A) suppressed conductivity and B) UV/Vis absorbance after postcolumn reaction with ODA. The bromate peak is baseline-resolved from chlorite on both detector channels; however, it shows a significantly enhanced response on the absorbance detector after PCR with ODA compared to the response obtained on the conductivity detector µs 0 Column: IonPac AG9-HC, AS9-HC Eluent: 9.0 mm Sodium carbonate Flow Rate: 1.3 ml/min Inj. Volume: 225 µl Detector: A) Suppressed conductivity, ASRS -ULTRA, AutoSuppression external water mode B) Absorbance, 450 nm Postcolumn Reagent: o -Dianisidine PCR Flow Rate: 0.7 ml/min Postcolumn Heater: 60 C Peaks: 1. Chlorite 15 µg/l (ppb) 2. Bromate Bromide 15 A 4. Chlorate Eluent Autosampler DX-500 PC10 PCR Reservoir AG9-HC Mixing Tee AS9-HC Absorbance Detector Electrolytic Supressor Conductivity Detector Knitted RX Coil PCH-2 60 C Effluent to waste Figure 1. IC system configuration for EPA Method AU B Figure 2. Separation of a low-ppb DBP anion standard using an IonPac AS9-HC column: A) suppressed conductivity detection and B) visible absorbance detection after PCR with o-dianisidine. 24 Determination of Inorganic Oxyhalide Disinfection Byproduct Anions and Bromide in Drinking Water Using Ion Chromatography with the Addition of a Postcolumn Reagent for Trace Bromate Analysis

25 Table 2 summarizes the calibration data and method detection limits (MDLs) obtained for the oxyhalide DBP anions and bromide using dual conductivity and PCR detection. The MDL for each analyte was established by making seven replicate injections of a reagent water blank fortified at a concentration of 3 to 5 times the estimated instrument detection limit. 2 The use of PCR addition and UV/Vis detection allows quantification of bromate down to 0.5 µg/l without compromising detection limits obtained with suppressed conductivity detection for the other anions of interest. 6 Note that the use of electronic smoothing (Olympic, 25 points, 5 sec, 1 iteration) of the UV/Vis signal improves the calculated MDL for bromate. 2 Figure 3 demonstrates the effect of smoothing on the performance of the PCR detection for a 1.0 µg/l bromate standard. No significant loss of peak response is observed after smoothing, although baseline noise is reduced by a factor of approximately 2, which results in a similar improvement in the detection limit (Table 2). Table 2 Linear Ranges and MDLs for Oxyhalides and Bromide Range r 2 MDL Standard Calculated MDL* Chlorite Bromate conductivity Bromide Chlorate Bromate UV/Vis (smoothed) Bromate UV/Vis (no smoothing) AU B AU 0 Column: IonPac AG9-HC, AS9-HC Eluent: 9.0 mm Sodium carbonate Flow Rate: 1.3 ml/min Inj. Volume: 225 µl Detection: A) Absorbance, 450 nm, no smoothing B) Absorbance, 450 nm, Olympic smoothing 25 points, 5 sec, 1 iteration Postcolumn Reagent: o -Dianisidine PCR Flow Rate: 0.7 ml/min Postcolumn Heater: 60 C A B Peaks: 1. Bromate-UV µg/l (ppb) A Figure 3. Effect of smoothing on bromate determination: A) unsmoothed data and B) smoothed data (Olympic, 25 points, 5 sec, 1 iteration). *MDL = (t) x (S) Where t = student s t value for a 99% confidence level and a standard deviation estimate with n - 1 degrees of freedom [t = 3.14 for seven replicates of the MDL standard], and S = standard deviation of the replicate analysis. 25 Determination of Inorganic Oxyhalide Disinfection Byproduct Anions and Bromide in Drinking Water Using Ion Chromatography with the Addition of a Postcolumn Reagent for Trace Bromate Analysis

26 Figures 4 7 illustrate the performance of the method for the determination of inorganic oxyhalide DBP anions and bromide in drinking and bottled water samples. Figure 4 shows the chromatograms from a direct injection of drinking water (from Sunnyvale, California) obtained by using dual A) suppressed conductivity and B) UV/ Vis absorbance after postcolumn reaction with ODA. Neither chlorite nor bromate are observed in the drinking water sample; however, bromide and chlorate (frequently observed as a disinfection byproduct from the use of hypochlorite) are well resolved from the sample matrix. Figure 5 shows the chromatograms of the same drinking water sample spiked with chlorite, bromate, bromide, and chlorate at levels of 108, 11.3, 36, and 72 µg/l, respectively. The chromatograms were obtained using, in series, dual A) suppressed conductivity and B) UV/Vis absorbance after postcolumn reaction with ODA. Quantitative recoveries were obtained for all anions, as shown in Table 3. The benefits of PCR with UV/Vis detection for bromate determination can clearly be seen in Figure 5B: bromate peak response is significantly enhanced compared to the response on the conductivity detector and no response is observed for the large peak from about 20 µg/l chloride that elutes immediately after bromate. The use of PCR with UV/Vis detection allows the quantification of bromate down to 0.5 µg/l in the presence of 200 mg/l chloride (a 400,000-fold excess) with no sample pretreatment Column: IonPac AG9-HC, AS9-HC Eluent: 9.0 mm Sodium carbonate Flow Rate: 1.3 ml/min Inj. Volume: 225 µl Detector: A) Suppressed conductivity, ASRS -ULTRA, external water B) Absorbance, 450 nm Postcolumn Reagent: o -Dianisidine PCR Flow Rate: 0.7 ml/min Postcolumn Heater: 60 C A Peaks: 3. Bromide 99.1 µg/l (ppb) 4. Chlorate 92.1 A Column: IonPac AG9-HC, AS9-HC Eluent: 9.0 mm Sodium carbonate Flow Rate: 1.3 ml/min Inj. Volume: 225 µl Detector: A) Suppressed conductivity, ASRS -ULTRA, external water B) Absorbance, 450 nm Postcolumn Reagent: o -Dianisidine PCR Flow Rate: 0.7 ml/min Postcolumn Heater: 60 C A Peaks: 1. Chlorite µg/l (ppb) 2. Bromate Bromide Chlorate B 5. Bromate-UV µs B AU 1.20 µs A B AU Figure 4. Determination of DBP anions in tap water: A) suppressed conductivity detection and B) visible absorbance detection after PCR with o-dianisidine Figure 5. Determination of DBP anions in spiked tap water: A) suppressed conductivity detection and B) visible absorbance detection after PCR with o-dianisidine Determination of Inorganic Oxyhalide Disinfection Byproduct Anions and Bromide in Drinking Water Using Ion Chromatography with the Addition of a Postcolumn Reagent for Trace Bromate Analysis

27 Figure 6 shows the chromatograms from a direct injection of bottled spring water obtained using, in series, dual A) suppressed conductivity and B) UV/ Vis absorbance after postcolumn reaction with ODA. In this instance, both bromate and bromide are observed in the bottled water sample. Bromate, which is formed during ozonation of source water containing bromide, is present at about 2 µg/l and can clearly be seen in the UV/ Vis chromatogram, although no peak is evident on the conductivity detector. Figure 7 shows the chromatograms of the same bottled water sample spiked with chlorite, bromate, bromide, and chlorate at levels of 126, 13.2, 42, and 84 µg/l, respectively. These chromatograms were obtained by using, in series, dual A) suppressed conductivity and B) UV/Vis absorbance after postcolumn reaction with ODA. Table 3 shows that quantitative recoveries were again obtained for all anions. Table 3 also shows the recoveries obtained for bromate spiked into the same drinking and bottled water samples at a lower concentration of 2.2 µg/l when using UV/Vis absorbance after postcolumn reaction with ODA. This method permits quantitative recoveries (80 120%) for bromate at levels down to 1 µg/l when using PCR and UV/Vis detection. Column: IonPac AG9-HC, AS9-HC Eluent: 9.0 mm Sodium carbonate Flow Rate: 1.3 ml/min Inj. Volume: 225 µl Detector: A) Suppressed conductivity, ASRS -ULTRA, external water B) Absorbance, 450 nm Postcolumn Reagent: o -Dianisidine PCR Flow Rate: 0.7 ml/min Postcolumn Heater: 60 C A Peaks: 3. Bromide µg/l (ppb) B 5. Bromate-UV 1.72 Column: IonPac AG9-HC, AS9-HC Eluent: 9.0 mm Sodium carbonate Flow Rate: 1.3 ml/min Inj. Volume: 225 µl Detector: A) Suppressed conductivity, ASRS -ULTRA, external water B) Absorbance, 450 nm Postcolumn Reagent: o -Dianisidine PCR Flow Rate: 0.7 ml/min Postcolumn Heater: 60 C A Peaks: 1. Chlorite µg/l (ppb) 2. Bromate Bromide Chlorate 89.3 B 5. Bromate-UV A 1.20 A µs µs B B AU AU Figure 6. Determination of DBP anions in bottled water: A) suppressed conductivity detection and B) visible absorbance detection after PCR with o-dianisidine. Figure 7. Determination of DBP anions in spiked bottled water: A) suppressed conductivity detection and B) visible absorbance detection after PCR with o-dianisidine. 27 Determination of Inorganic Oxyhalide Disinfection Byproduct Anions and Bromide in Drinking Water Using Ion Chromatography with the Addition of a Postcolumn Reagent for Trace Bromate Analysis

28 Table 3 Anion Recoveries for Spiked Water Samples Tap Water Bottled Water Anion* Amount Recovery Amount Recovery Added Added Chlorite % % Bromate conductivity % % Bromide % % Chlorate % % Bromate UV/Vis % % Bromate UV/Vis** % % *Data were obtained from multi-analyte spikes into tap and bottled water samples. **Bromate only (2.2 µg/l) was added to tap and bottled water samples to determine low level recovery for this anion using UV/Vis detection. Table 4 Bromate Recovery from Simulated Chlorine Dioxide Treated Waters (STW)* Spiked STW Fe (II)- Treated Amount Recovery Added Laboratory Fortified Matrix Fe (II)-Treated Amount Recovery Added STW 0 ND % STW % % STW % % STW % % STW % % * Chlorite present at 100 µg/l Removal of Chlorite Interference When chlorine dioxide is used to disinfect drinking water, the DBP anion chlorite is found in the finished drinking water. Chlorite, like bromate, reacts with o-dianisidine to form a complex that absorbs at 450 nm. High chlorite levels can interfere with quantification of low-level bromate. 2 One approach to minimize the interference from chlorite is to remove the chlorite by reduction with ferrous sulfate, as described in the Sample Preparation section. This treatment was evaluated by applying it to a series of simulated chlorine dioxide-treated tap waters, which had been spiked with varying levels of bromate, and the corresponding LFMs. The results, summarized in Table 4, show that acceptable recoveries of bromate are obtained after such treatment. This treatment approach is recommended when analysis of low-level bromate is required in chlorine dioxide-treated drinking waters. Summary The IC method described in this Application Note, which uses an IonPac AS9-HC column and suppressed conductivity detection, followed by postcolumn addition of o-dianisidine with UV/Vis detection specifically to enhance bromate response, allows the determination of all the key oxyhalide anions and bromide at low µg/l levels in drinking and bottled waters. The use of postcolumn addition and UV/Vis detection allows the quantification of bromate in the range of µg/l without compromising the suppressed conductivity detection of chlorite, bromide, and chlorate. Conductivity detection is recommended for the quantification of bromate in the range of µg/l. 28 Determination of Inorganic Oxyhalide Disinfection Byproduct Anions and Bromide in Drinking Water Using Ion Chromatography with the Addition of a Postcolumn Reagent for Trace Bromate Analysis

29 References 1. Joyce, R.J.; Dhillon, H.S. J. Chromatogr. A 1994, 671, Wagner, H.P.; Pepich, B.V.; Hautman, D.P.; Munch, D.J. J. Chromatogr. A 1999, 850, Kruithof, J.C.; Meijers, R.T. Water Supply 1995, 13, Fed. Regist., 1994, 59 (145), Fed. Regist., 1998, 63 (241), Jackson, L.K.; Joyce, R.J.; Laikhtman, M.; Jackson, P.E. J. Chromatogr. A 1998, 829, U.S. EPA Method 300.0, U.S. Environmental Protection Agency, Cincinnati, OH, U.S. EPA Method 300.1, U.S. Environmental Protection Agency, Cincinnati, OH, Walters, B.D.; Gordon, G.; Bubnis, B. Anal. Chem. 1997, 69, Weinberg, H.S.; Yamada, H. Anal. Chem. 1998, 70, Nowack, B.; Von Gunten, U. J. Chromatogr. A 1999, 849, Achilli, M.; Romele, L. J. Chromatogr. A 1999, 847, Charles, L.; Pepin, D. Anal. Chem. 1998, 70, Creed, J.T.; Magnuson, M.L.; Brockhoff, C.A. Environ. Sci. Technol. 1997, 31, U.S. EPA Method 317.0, U.S. Environmental Protection Agency, Cincinnati, OH, Method 4500-ClO 2.C; Greenberg, A.E.; Clesceri, L.S.; Eaton, A.D. (Eds.); Standard Methods for the Examination of Water and Wastewater, 18th ed., APHA: Washington, DC, Hautman, D.P.; Bolyard, M. J. Chromatogr. 1992, 602, 65. Suppliers Aldrich Chemical Co., P.O. Box 2060, Milwaukee, WI 53201, USA. Tel: Fluka, P.O. Box 2060, Milwaukee, WI 53201, USA. Tel: Pierce Chemical Co., 3747 North Meridian Road, P.O. Box 117, Rockford, IL 61105, USA. Tel: Sigma Chemical Co., P.O. Box 14508, St. Louis, MO 63178, USA. Tel: SPEX CertiPrep, Inc., 203 Norcross Ave., Metuchen, NJ 08840, USA. Tel.: VWR Scientific Products, 3745 Bayshore Blvd., Brisbane, CA 94005, USA. Tel: Determination of Inorganic Oxyhalide Disinfection Byproduct Anions and Bromide in Drinking Water Using Ion Chromatography with the Addition of a Postcolumn Reagent for Trace Bromate Analysis

30 Application Note 208 Determination of Bromate in Bottled Mineral Water Using the CRD 300 Carbonate Removal Device INTRODUCTION Drinking and bottled waters are commonly disinfected with ozone. Ozone is highly effective and, unlike many other disinfectants, does not remain in the water or change its taste. Unfortunately, when bromide is present in water, it is converted to bromate by the ozone treatment. Bromate is recognized as a potential human carcinogen, which has led to the regulation of its concentration in drinking and bottled water. Major regulatory bodies worldwide (e.g., U.S. EPA and the European Commission) have set a maximum allowable bromate concentration in drinking water of 10 µg/l. 1 In Europe, the limit was lowered to 3 µg/l for bottled natural mineral and spring waters disinfected by ozonation. 2 Over the past two decades, Dionex has led the effort in developing sensitive and robust ion chromatography (IC) methods for determining bromate and other oxyhalides (e.g., chlorite and chlorate). U.S. EPA Method (B) and (B) used the IonPac AS9-SC and IonPac AS9-HC columns, respectively, along with suppressed conductivity detection for bromate, chlorite, and chlorate determinations in drinking water. In 1997, Dionex introduced the AS9-HC column to allow the direct injection of 250 µl of drinking water to easily meet the 10 µg/l regulatory requirement. This method was documented in Dionex Application Note 81 (AN 81). 3 Since then, Dionex has developed a number of products and techniques, and worked with regulatory agencies and international standards organizations to improve the sensitivity and ruggedness of bromate determinations as well as the types of samples that can be directly injected. Dionex products were instrumental in the development of the postcolumn derivatization techniques in U.S. EPA methods and These methods used the AS9-HC and Dionex suppression technology for conductivity detection of oxyhalides combined with postcolumn addition and absorbance detection for enhanced determination of bromate. EPA Methods and are documented in AN 136 and AN ,5 To improve the sensitivity for bromate using direct injection, Dionex developed the IonPac AS19 column. This column was designed for use with hydroxide eluents rather than the carbonate eluents used with the AS9-HC. Hydroxide eluents offer improved sensitivity for suppressed conductivity detection as compared to carbonate eluents. This improved sensitivity was documented in AN Hydroxide eluents are also advantageous because they can be generated easily using an eluent generator as part of a Reagent-Free IC (RFIC ) system. RFIC systems improve reproducibility and simplify analysis. The AS19 separation can also replace the AS9-HC separation in EPA Methods and 326.0, which is documented in AN 168 and AN ,8 The AS19 was also used with an isocratic hydroxide eluent rather than the typical gradient for analysis of drinking water for bromate. 9 This method, presented in Application Update 154 (AU 154), cannot determine all the common inorganic anions in a single injection like the gradient method in AN 167. For determination of sub-µg/l concentrations of bromate 30 Determination of Bromate in Bottled Mineral Water Using the CRD 300 Carbonate Removal Device

31 in drinking water and higher ionic strength matrices without postcolumn derivatization, Dionex developed a two-dimensional IC technique (AN 187) that uses an AS19 column in the first dimension, and an AS24 column, developed specifically for determining haloacetic acids and bromate by IC-MS and IC-MS/MS, in the second dimension. 10 Dionex AN 184 showed that the AS19 method in AN 167 could be used to meet the 3 µg/l European limit for bromate in natural mineral and spring waters disinfected by ozonation. 11 The same application note compared the AS19 chromatography to chromatography with the AS23, a column that uses carbonate eluents and was designed to replace the AS9-HC. The AS23 has a higher capacity than the AS9-HC, and a different selectivity for the carbonate ion so that it is less likely to interfere with bromate determinations. AN 184 showed that poorer sensitivity associated with using carbonate eluents when compared to hydroxide eluents made the AS23 performance inferior to that of the AS19. The present application note describes the use of a carbonate removal device, the CRD 300, to remove the majority of carbonate from the eluent and allow hydroxide-like performance and detection sensitivity. This device was used with the IonPac AS23 to determine bromate in a bottled mineral water samples. Detection sensitivity when using the CRD 300 was improved compared to chromatography without the CRD 300. Scientists responsible for water analysis can choose the column and eluent chemistry that best meets their needs to reliably determine bromate at concentrations below the common 10 µg/l regulatory limit. EQUIPMENT Dionex ICS-2000 Reagent-Free Ion Chromatography System* equipped with the following for carbonate/ bicarbonate eluent generation: EluGen EGC II K 2 CO 3 cartridge (P/N ) EPM Electrolytic ph Modifier (P/N ) EGC Carbonate Mixer (P/N ) CRD 300 Carbonate Removal Device (4 mm) with VC Vacuum Pump (P/N ) Chromeleon 6.8 Chromatography Management Software REAGENTS AND STANDARDS Deionized water, type I reagent grade, 18 MΩ-cm resistivity or better Sodium chlorite, 80% (NaClO 2, Fluka) Potassium bromate (KBrO 3, Fluka) Sodium chlorate (NaClO 3, Fluka) Individual stock standards of fluoride, chloride, and sulfate, 1000 mg/l each (Merck) PREPARATION OF SOLUTIONS AND REAGENTS Carbonate Eluent Generation The Eluent Generator (EG) produces the eluent using the EluGen EGC II K 2 CO 3 cartridge, Electrolytic ph Modifier, EGC Carbonate Mixer, and deionized water supplied by the pump. The eluent concentration is controlled by the Chromeleon software. Backpressure tubing must be added to achieve psi backpressure that will allow the EG degasser to function properly. See the ICS-2000 Operator s Manual Section 2.4.4, "Eluent Generator" for instructions on adding backpressure. To set up the EGC II K2CO3, see the EGC II K2CO3 cartridge, Electrolytic ph Modifier, and EGC Carbonate Mixer Product Manual (Doc. No ) for more information. Manual Eluent Preparation From Eluent Concentrate Prepare 1 L of eluent by adding 10 ml of the Dionex IonPac AS23 Eluent Concentrate (P/N ) to a 1 L volumetric flask. Bring to volume with DI water and mix thoroughly. *This application can be run on any Dionex system equipped for carbonate/bicarbonate eluent generation. Alternately, this application can be run with a manually prepared carbonate/bicarbonate eluent. 31 Determination of Bromate in Bottled Mineral Water Using the CRD 300 Carbonate Removal Device

32 From Manually Prepared Stock Solutions Stock Carbonate/Bicarbonate Eluent Preparation 1.0 M Na 2 CO 3 and 1.0 M NaHCO 3 Weigh g sodium carbonate and g sodium bicarbonate into separate 100 ml volumetric flasks. Bring each to volume with DI water. IonPac AS23 Eluent (4.5 mm Na 2 CO 3 /0.8 mm NaHCO 3 ) For 1L, prepare by adding 4.5 ml of 1.0 M Na 2 CO 3 and 0.8 ml of 1.0 M NaHCO 3 to a 1L volumetric flask, bring to volume with DI water, and mix thoroughly. Stock Standard Solutions Prepare 1000 mg/l stock standard solutions of fluoride, chloride, sulfate, chlorite, bromate, and chlorate by weighing g, g, g, g, g, and g, respectively, into separate 100 ml volumetric flasks. Bring each to volume with DI water. Secondary Standards The stock standards are used to prepare the 1000 µg/l secondary standards of chlorite, bromate, and chlorate. Take a defined volume of the stock standard and dilute it 1 to 1000 with DI water (e.g., dilute 100 µl to 100 ml in a 100 ml volumetric flask). Use these standards to prepare the working standards and to spike the bottled mineral water sample. Working Standards Prepare the standards for calibration and MDL studies by mixing defined volumes of the 1000 mg/l stock stadard solutions of fluoride, chloride, and sulfate and the 1000 µg/l secondary standards of chlorite, bromate, and chlorate. For example, to prepare the working standard containing 0.5 mg/l fluoride, 50 mg/l chloride, 100 mg/l sulfate, and 40 µg/l of each of the oxyhalides, add 0.05 ml of the fluoride stock standard, 5 ml of the chloride stock standard, 10 ml of the sulfate stock standard, and 4 ml of each oxyhalide secondary standard to a 100 ml volumetric flask and bring to volume. Sample The bottled mineral water sample was purchased from a local market in Bangkok, Thailand and was bottled at its source in the mountains of Thailand. The label reported the presence of fluoride, chloride, sulfate, and bicarbonate, but not their concentrations. CRD 300 in Vacuum Mode Setup The CRD 300 in vacuum mode uses a vacuum pump to evacuate the regenerant chamber of the CRD 300 so that CO 2 gas is literally sucked out of the eluent. A bleed tube feeds a trickle of fresh air into the regenerant chamber to constantly sweep out the CO 2 gas. To operate the CRD 300 in vacuum mode, mount the CRD 300 directly on top of the suppressor and plumb the eluent from the Eluent Out of the suppressor to the Eluent In of the CRD 300. The Eluent Out of the CRD 300 is connected to the conductivity cell In and conductivity cell Out goes to waste if the system is running in external water mode. If the system is operated in recycle mode, connect conductivity cell Out to the suppressor Regen In. Connect the vacuum tubing to the vacuum port of the vacuum pump and to the ballast bottle. Connect a length of 1/8 Teflon tubing from the ballast bottle to the Regen Out of the CRD 300. Make sure the third port on the ballast bottle is closed and air tight. Connect 15 cm of red (0.005 i.d.) PEEK tubing to the Regen In of the CRD 300; this is the air bleed assembly. Begin eluent flow before beginning vacuum operation. When eluent flow is established, turn on the vacuum pump. The background conductivity should drop almost immediately. When the eluent pump is turned off, immediately turn off the vacuum pump. Avoid operating the vacuum pump while eluent flow is stopped. A TTL can be wired to automate stopping the vacuum pump. 32 Determination of Bromate in Bottled Mineral Water Using the CRD 300 Carbonate Removal Device

33 CONDITIONS Condition A (Eluent Generation and CRD 300) Column: IonPac AS23 (4 250 mm) (P/N ) IonPac AG23 (4 50 mm) (P/N ) Eluent: EGC II K 2 CO 3 (P/N ) Flow Rate: Inj. Volume: 250 µl Temperature: 30 C Suppressor: Background: < 1.5 µs Noise: EPM (P/N ) 4.5 mm K 2 CO 3 /0.8 mm KHCO ml/min Suppressed conductivity, ASRS 300, 4 mm (P/N ), external water mode, 25 ma CRD 300, 4 mm, (P/N ) vacuum mode ~ 0.3 ns Back Pressure: ~2200 psi Condition B (Manual Eluent Preparation and no CRD 300) Column: IonPac AS23 (4 250 mm) (P/N ) IonPac AG23 (4 50 mm) (P/N ) Eluent: 4.5 mm Na 2 CO 3 /0.8 mm NaHCO 3 Flow Rate: 1.0 ml/min Inj. Volume: 250 µl Column Temp: 30 C Suppressor: Background: µs Noise: Suppressed conductivity, ASRS 300, 4 mm (P/N ), external water mode, 25 ma ~ 3.0 ns Back Pressure: ~1800 psi RESULTS AND DISCUSSION Chromatography Bromate, chlorite, and chlorate were resolved from seven common inorganic anions using an IonPac AS23 column under its recommended eluent conditions (4.5 mm Na 2 CO 3 /0.8 mm NaHCO 3 ). Chromatogram B in Figure 1 shows this separation. The background conductivity after suppression using the carbonate eluent is between 18 and 19 µs. The higher the background, the higher the noise, and this results in a lower signal-to-noise ratio (i.e., lower sensitivity). The background of the suppressed hydroxide eluent used for the IonPac AS19 column is < 1 µs. In order for the carbonate eluent system of the AS23 to approach the detection limits delivered by the hydroxide eluent system of the AS19, the background must be reduced. The CRD 300 was designed to remove carbonate from the eluent (after suppression) and thereby reduce the background to improve detection limits. Chromatogram A shows the same AS23 separation as B using a CRD 300. Note that the background has been reduced to about 1 µs, the injection dip at about 2 min is greatly reduced in size, and there is a noticeable improvement in analyte sensitivity. Throughout this application note, we compare the determination of bromate, chlorite, and chlorate with the AS23 and suppressed conductivity, both with and without the CRD µs A B Column: IonPac AS23 (4 250 mm) IonPac AG23 (4 50 mm) Eluent: A) EGC II K 2CO 3/EPM 4.5 mm K 2 CO 3 /0.8 mm KHCO 3 B) 4.5 mm Na 2CO 3/0.8 mm NaHCO 3 Temperature: 30 C Flow Rate: 1.0 ml/min Inj. Volume: 250 µl Detection: A) Suppressed conductivity, ASRS 300, external water mode with CRD 300 in vacuum mode B) Suppressed conductivity, ASRS 300, external water mode Suppressor Current: 25 ma µs Peaks: 1. Fluoride 0.1 mg/l 2. Chlorite Bromate Chloride Nitrite Chlorate Bromide Nitrate Phosphate Sulfate Figure 1. Chromatography of a mixed anion standard A) with a CRD 300 and electrolytically prepared eluent, and B) without a CRD 300 and with manually prepared eluent. 33 Determination of Bromate in Bottled Mineral Water Using the CRD 300 Carbonate Removal Device

34 Figure 2 shows single injections from the MDL determinations of bromate, chlorite, and chlorate with and without the CRD 300. Fluoride, (0.5 mg/l), chloride (50 mg/l), and sulfate (100 mg/l) were added to the MDL standards to simulate the ionic strength of bottled water samples. Due to the higher background and noise of the system without the CRD 300 (Chromatogram B, Figure 2), higher analyte concentrations were used for the MDL test compared to the system with the CRD 300. Table 1 shows the results of the MDL determination. For all three oxyhalide analytes, the MDL is lower for the system with the CRD 300. The MDL values without the CRD 300 are similar to those determined with the AS23 in AN 184. The values when using the CRD 300, though lower than without, are not as low as those determined with the AS19 and hydroxide eluent in AN A Column: IonPac AS23 (4 250 mm) IonPac AG23 (4 50 mm) 4 Eluent: A) EGC II K 2 CO 3 /EPM mm K 2 CO 3 /0.8 mm KHCO 3 B) 4.5 mm Na 1 2 CO 3 /0.8 mm NaHCO 3 Temperature: 30 C µs 3 Flow Rate: 1.0 ml/min 2 5 Inj. Volume: 250 µl Detection: A) Suppressed conductivity, ASRS 300, external water mode with CRD 300 in vacuum mode 0.2 B) Suppressed conductivity, ASRS 300, external water mode 0.5 B Suppressor 4 Current: 25 ma 6 Peaks: A B µs 1 1. Fluoride mg/l 3 2. Chlorite 4 8 µg/l Bromate 5 10 µg/l 4. Chloride mg/l 5. Chlorate 4 8 µg/l 6. Sulfate mg/l Figure 2. Example chromatograms from the MDL determination A) with a CRD 300, and B) without a CRD 300. Table 1. MDL Determinations of Chlorite, Bromate, and Chlorate with and without a CRD 300 Height (µs) With CRD 300 Without CRD 300 Chlorite Bromate Chlorate Chlorite Bromate Chlorate Injection No. 4 µg/l 5 µg/l 4 µg/l 8 µg/l 10 µg/l 8 µg/l Average RSD MDL Determination of Bromate in Bottled Mineral Water Using the CRD 300 Carbonate Removal Device

35 Another calibration was performed for both systems using consistent concentrations of fluoride, chloride, and sulfate (0.5 mg/l, 2 mg/l, and 10 mg/l, respectively) in standards with three levels of chlorite, bromate, and chlorate concentrations; 10, 20, and 40 µg/l. Overlays of three calibration standards are shown in Figure 3 and the results are in Table 2. The calibration data are equivalent. Both systems were used to analyze a bottled mineral water sample from the mountains of Thailand. Figure 4 shows the analysis of this sample and Table 3 reports the results of the analysis. The sample had just over 10 µg/l bromate and 1 2 µg/l chlorate, suggesting a second disinfection process besides ozonation was used. Due to the noise of the system without the CRD 300, the chlorate peak could not be identified with confidence. To evaluate accuracy, known amounts of bromate, chlorite, and chlorate were spiked into the bottled mineral water sample. Figure 5 shows the chromatography from this study and Table 4 shows that all analytes were recovered at >85%. In this experiment, the recovery was better for the system with the CRD µs µs A B Column: IonPac AS23 (4 250 mm) IonPac AG23 (4 50 mm) Eluent: A) EGC II K 2 CO 3 /EPM 4.5 mm K 2 CO 3 /0.8 mm KHCO 3 B) 4.5 mm Na 2 CO 3 /0.8 mm NaHCO 3 Temperature: 30 C Flow Rate: 1.0 ml/min Inj. Volume: 250 µl Detection: A) Suppressed conductivity, ASRS 300, external water mode with CRD 300 in vacuum mode B) Suppressed conductivity, ASRS 300, external water mode Suppressor Current: 25 ma Peaks: 1. Fluoride 0.5 mg/l 2. Chlorite 10, 20, 40 µg/l 3. Bromate 10, 20, 40 µg/l 4. Chloride 2.0 mg/l 5. Chlorate 10, 20, 40 µg/l 6. Sulfate 10.0 mg/l Figure 3. Overlay of chromatograms of three concentration levels of chlorite, bromate, and chlorate in a mixed anion standard A) with a CRD 300, and B) without a CRD Table 2. Chromeleon Calibration Report for Chlorite, Bromate, and Chlorate with and without a CRD 300 Peak Name Points R 2 (%) With CRD 300 Without CRD 300 Chlorite Bromate Chlorate Determination of Bromate in Bottled Mineral Water Using the CRD 300 Carbonate Removal Device

36 0.10 A µs B µs 2 3 Table 3. Determination of Bromate and Chlorate in a Bottled Mineral Water Sample with and without a CRD 300 Injection No With CRD 300 Column: IonPac AS23 (4 250 mm) IonPac AG23 (4 50 mm) Eluent: A) EGC II K 2 CO 3 /EPM 4.5 mm K 2 CO 3 /0.8 mm KHCO 3 B) 4.5 mm Na 2 CO 3 /0.8 mm NaHCO 3 Temperature: 30 C Flow Rate: 1.0 ml/min Inj. Volume: 250 µl Detection: A) Suppressed conductivity, ASRS 300, external water mode with CRD 300 in vacuum mode B) Suppressed conductivity, ASRS 300, external water mode Suppressor Current: 25 ma Sample: Bottled Mineral Water Peaks: A B 1. Fluoride 2. Formate 3. Bromate µg/l 4. Chloride 5. Chlorate 1.56 ND µg/l 6. Nitrate 7. Carbonate 8. Phosphate 9. Sulfate 10. Unknown Figure 4. Chromatography of a bottled mineral water sample A) with a CRD 300, and B) without a CRD Without CRD 300 Bromate Chlorate Bromate Chlorate ND ND ND ND ND Average RSD A µs B µs Column: IonPac AS23 (4 250 mm) IonPac AG23 (4 50 mm) Eluent: A) EGC II K 2 CO 3 /EPM 4.5 mm K 2 CO 3 /0.8 mm KHCO 3 B) 4.5 mm Na 2 CO 3 /0.8 mm NaHCO 3 Temperature: 30 C Flow Rate: 1.0 ml/min Inj. Volume: 250 µl Detection: A) Suppressed conductivity, ASRS 300, external water mode with CRD 300 in vacuum mode B) Suppressed conductivity, ASRS 300, external water mode Suppressor Current: 25 ma Sample: Bottled mineral water spiked with chlorite, bromate, and chlorate Peaks: A B 1. Fluoride 2. Formate 3. Chlorite µg/l 4. Bromate µg/l 5. Chloride 6. Chlorate µg/l 7. Unknown-1 8. Carbonate 9. Phosphate 10. Sulfate 11. Unknown Figure 5. Chromatography of a bottled mineral water sample spiked with chlorite, bromate, and chlorate (10 µg/l each) A) with a CRD 300, and B) without a CRD 300. Table 4. MDL Determinations of Chlorite, Bromate, and Chlorate with and without a CRD 300 With CRD 300 Without CRD 300 Chlorite Bromate Chlorate Chlorite Bromate Chlorate Sample ND a ND a 5.74 ND a Spike Measured b Amount RSD Recovery(%) a ND = Not Detected; b The average of five injections 36 Determination of Bromate in Bottled Mineral Water Using the CRD 300 Carbonate Removal Device

37 Summary This application note shows that using the CRD 300 with the IonPac AS23, bromate can be determined in bottled mineral water at concentrations < 5 µg/l. The method sensitivity for bromate and other oxyhalides approaches that of the hydroxide eluent system featured in Dionex Application Note 184. References 1. US EPA. National Primary Drinking Water Regulations: Disinfectants and Disinfection Byproducts. Fed. Regist. 1998, 63 (241), European Parliament and Council Directive No. 2003/40/EC, Establishing the List, Concentration Limits and Labeling Requirements for the Constituents of Natural Mineral Waters and the Conditions for Using Ozone-Enriched Air for the Treatment of Natural Mineral Waters and Spring Waters, Dionex Corporation. Ion Chromatographic Determination of Oxyhalides and Bromide at Trace Level Concentrations in Drinking Water Using Direct Injection, Application Note 81, LPN Sunnyvale, CA, Dionex Corporation. Determination of Inorganic Oxyhalide Disinfection Byproduct Anions and Bromide in Drinking Water Using Ion Chromatography with the Addition of a Postcolumn Reagent for Trace Bromate Analysis, Application Note 136, LPN Sunnyvale, CA, Dionex Corporation. Determination of Chlorite, Bromate, Bromide, and Chlorate in Drinking Water by Ion Chromatography with an On-Line-Generated Postcolumn Reagent for Sub-µg/L Bromate Analysis, Application Note 149, LPN Sunnyvale, CA, Dionex Corporation. Determination of Trace Concentrations of Oxyhalides and Bromide in Municipal and Bottled Waters Using a Hydroxide- Selective Column with a Reagent-Free Ion Chromatography System, Application Note 167, LPN Sunnyvale, CA Dionex Corporation. Determination of Trace Concentrations of Disinfection By-Product Anions and Bromide in Drinking Water Using Reagent- Free Ion Chromatography Followed by Postcolumn Addition of o-dianisidine for Trace Bromate Analysis, Application Note 168, LPN Sunnyvale, CA, Dionex Corporation. Determination of Disinfection By-Product Anions and Bromide in Drinking Water Using a Reagent-Free Ion Chromatography System Followed by Postcolumn Addition of an Acidified On- Line Generated Reagent for Trace Bromate Analysis, Application Note 171, LPN Sunnyvale, CA, Dionex Corporation. Determination of Bromate in Drinking and Natural Mineral Water by Isocratic Ion chromatography with a Hydroxide Eluent, Application Update 154, LPN Sunnyvale, CA, Dionex Corporation. Determination of Sub-µg/L Bromate in Municipal and Natural Mineral Waters Using Preconcentration with Two-Dimensional Ion Chromatography and Suppressed Conductivity Detection, Application Note 187, LPN Sunnyvale, CA, Dionex Corporation. Determination of Chlorite, Bromate, and Chlorate in Bottled Natural Mineral Waters, Application Note 184, LPN Sunnyvale, CA, Determination of Bromate in Bottled Mineral Water Using the CRD 300 Carbonate Removal Device

38 Application Note 184 Determination of Trace Concentrations of Chlorite, Bromate, and Chlorate in Bottled Natural Mineral Waters INTRODUCTION Bottled water has been one of the fastest growing beverage markets in the last five to ten years. Global consumption approached 41 billion gallons in 2004, an increase of 6.5% from The bottled water industry markets to health conscious consumers as an alternative not only to tap water, but also to carbonated soft drinks and juice drinks. 1 Regardless of whether the water is delivered from a local municipality or is prepackaged in a bottle, the consumption of safe and reliable drinking water is essential to maintain a healthy lifestyle. Bottled water must be disinfected to remove pathogenic microorganisms and ensure it is safe for human consumption. Water companies prefer ozone as a disinfectant because it is one of the most effective treatments available, it does not leave a taste, and there is no residual disinfectant in the bottled water. 2,3 Some bottlers, however, use ultraviolet light or chlorine dioxide as alternative treatment methods. 2 Reactions between disinfectants and natural organic and inorganic matter in the source water can result in the production of undesirable disinfection byproducts (DBPs), such as chlorite, bromate, and trihalomethanes, that are potentially harmful to humans. 4 Bromate, for example, can be formed by ozonation of water containing naturally occurring bromide, or may be present as an impurity in sodium hypochlorite used for treatment. 5 Results from toxicological studies led the International Agency for Research on Cancer to conclude that bromate is a potential human carcinogen, even at low μg/l (ppb) concentrations. 6 The World Health Organization (WHO) estimated excess lifetime cancer risks of 10-4, 10-5, and 10-6 for drinking water containing bromate at 20, 2, and 0.2 μg/l, respectively. 9 The U.S. EPA, 7 European Commission, 8 and the WHO 9 set a maximum permissible limit of 10 μg/l bromate in tap water. The U.S. FDA 10 adopted the same regulatory limit for bottled water. In Europe, natural mineral waters and spring waters treated by ozonation have a maximum permissible limit of 3 μg/l bromate. 11 Traditionally, ion chromatography (IC) with suppressed conductivity detection has been used for determination of bromate and other DBPs in drinking water, as described in EPA Method This method describes the use of a high-capacity IonPac AS9-HC column with a carbonate eluent and large loop injection to achieve a method detection limit (MDL) of 1.4 μg/l bromate. In early 2006, the U.S. EPA enacted stage 2 of the disinfectants/disinfection byproducts (D/DBP) rule, maintaining the maximum permissible limit for bromate but adding three additional analytical methods to further improve the selectivity and sensitivity for bromate. 13 U.S. EPA Methods and combine suppressed conductivity detection and absorbance detection after postcolumn addition to achieve bromate MDLs less than 0.2 μg/l. 14,15 IC coupled to inductively coupled plasma mass spectrometry has also been demonstrated for the determination of low concentrations of bromate in environmental waters, permitting a bromate MDL of 0.3 μg/l Determination of Trace Concentrations of Chlorate, Bromate, and Chlorite in Bottled Natural Mineral Waters

39 A high-capacity IonPac AS19 column with an electrolytically generated hydroxide eluent, large loop injection, and suppressed conductivity detection can achieve a calculated bromate MDL of 0.34 μg/l. 17 Absorbance detection after postcolumn addition can reduce this MDL to less than 0.2 μg/l, using EPA Methods and ,19 In this application note, we compare the IonPac AS19 using an electrolytically generated hydroxide eluent to the IonPac AS23 column using an electrolytically generated carbonate/bicarbonate eluent for the determination of chlorite, bromate, and chlorate in natural mineral waters. We compare the linearity, method detection limits, precisions, and recovery for three mineral waters obtained from three European countries to determine whether these columns have the sensitivity required to meet current EPA and EU requirements. EQUIPMENT A Dionex ICS-2000 Reagent-Free Ion Chromatography (RFIC ) system was used in this work. The ICS-2000 is an integrated ion chromatograph and consists of: Eluent generator Pump with in-line vacuum degas Column heater Hydroxide system: EluGen EGC II KOH cartridge (Dionex P/N ) CR-ATC (Dionex P/N ) Carbonate system: EluGen EGC II K 2 CO 3 cartridge (Dionex P/N ) EPM Electrolytic ph Modifier to generate the carbonate/bicarbonate eluent (Dionex P/N ) EGC Carbonate Mixer (Dionex P/N ) Two 4-L plastic bottle assemblies (for external water mode of suppression) AS Autosampler Chromeleon Chromatography Management Software REAGENTS AND STANDARDS Deionized water, type I reagent grade, 18 MΩ-cm resistivity or better Sodium chlorite (NaClO 2, Fluka 71388, 80% pure) Sodium bromate (NaBrO 3, EM SX ) Sodium chlorate (NaClO 3, Fluka 71370) CONDITIONS Columns: Eluent: (A) IonPac AS19 Analytical, mm (Dionex P/N ) IonPac AG19 Guard, 4 50 mm (Dionex P/N ) (B) IonPac AS23 Analytical, mm (Dionex P/N ) IonPac AG23 Guard, 4 50 mm (Dionex P/N ) (A) 10 mm KOH from 0 10 min, mm from min, 45 mm from min* (B) 4.5 mm K 2 CO 3 /0.8 mm KHCO 3 Eluent Source: (A) EGC II KOH with CR-ATC (B) EGC II K 2 CO 3 with EPM Flow Rate: 1.0 ml/min Temperature: 30 C Injection: 250 μl Detection: (A) Suppressed conductivity, ASRS ULTRA II, 4 mm (Dionex P/N ) AutoSuppression recycle mode 130 ma current (B) Suppressed conductivity, ASRS ULTRA II, 4 mm AutoSuppression external water mode 25 ma current CRD: (A) 4-mm format (P/N ) Background Conductance: System Backpressure: Run Time: (A) <1 μs (B) μs ~2200 psi 30 min *Method returns to 10 mm KOH for 3 min prior to injection. PREPARATION OF SOLUTIONS AND REAGENTS Eluent Solution for the AS23 Column 4.5 mm Carbonate/0.8 mm Bicarbonate Generate the carbonate/bicarbonate eluent on-line by pumping high quality deionized water (18 MΩ-cm resistivity or better) through the EluGen EGC II K 2 CO 3 Cartridge and EPM. Chromeleon will track the amount of eluent used and calculate the remaining lifetime. Alternatively, prepare the eluent solution by adding 10 ml of the AS23 Eluent Concentrate (Dionex 39 Determination of Trace Concentrations of Chlorate, Bromate, and Chlorite in Bottled Natural Mineral Waters

40 P/N ) to a 1-L volumetric flask containing approximately 700 ml of degassed deionized water. Bring to volume and mix thoroughly. The 0.45 M sodium carbonate/0.08 M sodium bicarbonate concentrate can also be prepared from the salts by combining 47.7 g sodium carbonate (MW=106 g/mole) and 6.72 g sodium bicarbonate (MW=84 g/mole) in a 1-L volumetric flask containing approximately 700 ml of degassed deionized water. Bring to volume and mix thoroughly. Stock Standard Solutions Prepare 1000 mg/l stock standard solutions of chlorite, bromate, and chlorate by dissolving g, g, and g, respectively, of the corresponding sodium salts in separate 100 ml volumetric flasks of DI water. Calibration Standard Solutions Prepare a secondary stock solution containing 1 mg/l each of chlorite and chlorate and a separate secondary stock solution containing 1 mg/l bromate by performing the appropriate dilutions of the 1000 mg/l stock standards. Calibration standards can then be prepared from the secondary solutions using the appropriate dilutions. Dilute working standards should be prepared monthly, except those that contain chlorite, which must be prepared every two weeks, or sooner if evidence of degradation is indicated by repeated QC failures. Concentration ranges used in this application note are shown in Table 1. Sample PREPARATION For the present analysis, mineral waters B and C were degassed for min under vacuum due to an excess amount of bicarbonate in the samples. Increased amounts of bicarbonate in the sample can produce shifts in retention times as shown in Figure 1 traces A and B. In addition, due to the presence of significantly high concentrations of sulfate in mineral water C, the sample was diluted 1:5 with DI water prior to analysis. Column: IonPac AG19, AS19, 4 mm Eluent: 10 mm KOH 0-10 min, mm min, 45 mm min Eluent Source: EGC-KOH with CR-ATC Temperature: 30 C Flow Rate: 1.0 ml/min Inj. Volume: 250 µl Detection: Suppressed conductivity ASRS ULTRA II, recycle mode 2 µs Peaks: Fluoride 2. Chloride 3. Bromate 4. Nitrate 5. Carbonate 6. Sulfate 7. Unknown 6 B A Figure 1. Comparison of mineral water B A) before vacuum degas and B) after vacuum degas. Table 1. Calibration Data, Retention Time Precisions, Peak Area Precisions, and Method Detection Limits For DBP Anions IonPac AS19 Column Analyte Range (µg/l Linearity r 2 Retention Time* RSD (%) Peak Area RSD (%) MDL Standard Calculated MDL Chlorite Bromate Chlorate IonPac AS23 Column Chlorite Bromate Chlorate a RSD= relative standard deviation, n = 10 for a standard consisting of 10 ppb bromate and 20 ppb each of chlorite and chlorate. 40 Determination of Trace Concentrations of Chlorate, Bromate, and Chlorite in Bottled Natural Mineral Waters

41 RESULTS AND DISCUSSION The IonPac AS23 is a high-capacity anion-exchange column specifically designed to be used with carbonate /bicarbonate eluent for the determination of the trace DBPs, chlorite, bromate, and chlorate, together with common inorganic anions, including bromide (precursor to bromate), in drinking waters. To simplify the method and avoid manual eluent preparation, this column can be used with electrolytically generated potassium carbonate that is modified by an Electrolytic ph Modifier (EPM) to automatically generate the carbonate/bicarbonate eluent that is required for analyte separation. The IonPac AS23 column was developed using a unique polymer technology to achieve a capacity of 320 μeq/column, higher than the IonPac AS9-HC column (190 μeq/column) described in EPA Method The combination of an optimized selectivity for DBP anions, high anion exchange capacity, and improved selectivity of carbonate from inorganic anions and oxyhalides, makes this column an ideal replacement for the AS9-HC column. In this application, we compare the IonPac AS23 column to the hydroxide-selective IonPac AS19 column for the determination of trace DBP anions in natural mineral waters. Figure 2 compares the separation for chlorite, bromate, and chlorate on the IonPac AS19 and AS23 columns. As shown, both columns provide good selectivity for the target DBP anions. The linear calibration ranges, MDLs, and quality control standard (QCS) performances were evaluated for the hydroxide and carbonate eluent systems. The hydroxide eluent system was calibrated using four increasing concentrations of chlorite and chlorate (2-50 μg/l) and five increasing concentrations of bromate (1-25 μg/l). For the carbonate-based system, chlorite and chlorate were calibrated from μg/l whereas bromate was calibrated from 5-25 μg/l using three different concentrations. Each system produced a linear response in its respective range with a correlation coefficient greater than The improved sensitivity of the hydroxide eluent system, however, allowed a lower minimum reporting limit (MRL) than the carbonate-based system. The MDLs for the target DBPs were determined for each system by performing seven replicate injections of reagent water fortified with the calibration standards at concentrations of three to five times the estimated instrument detection limits. Column: Eluent: A) IonPac AG19, AS19, 4 mm B) IonPac AG23, AS23, 4 mm A) 10 mm KOH 0-10 min, mm min, 45 mm min B) 4.5 mm potassium carbonate/ 0.8 mm potassium bicarbonate Eluent Source: A) EGC-KOH with CR-ATC B) EGC-K 2 CO 3 + EPM Temperature: 30 C Flow Rate: 1.0 ml/min Inj. Volume: 250 µl µs A B µs Detection: Suppressed conductivity ASRS ULTRA II, recycle mode A) Recycle mode B) External mode Peaks: 1. Chlorite 10 µg/l 2. Bromate 5 3. Chlorate Figure 2. Separation of disinfection byproducts using the A) IonPac AS19 column and B) IonPac AS23 column. Table 1 compares the calibration data, retention time and peak area precisions for a QCS, and MDLs for the IonPac AS19 with an electrolytically generated hydroxide eluent to the IonPac AS23 with an electrolytically generated carbonate/bicarbonate eluent. The calculated MDL of bromate with the IonPac AS19 column was 0.31 μg/l compared to 1.63 μg/l using the IonPac AS23 column. This demonstrates that hydroxide eluents improve the sensitivity for bromate compared to carbonate-based eluents and are therefore more suitable to meet the current European regulatory requirement of 3 μg/l bromate in natural mineral waters. Either the AS19 or AS23 based IC systems are capable of measuring the 10 μg/l requirement of bromate for tap water or U.S. bottled water according to the regulations established by the U.S. EPA, U.S. FDA, WHO, and European Commission. In the U.S., mineral water is defined as water that contains no less than 250 ppm total dissolved solids (TDS) and that originates from a geologically and physically protected underground water source. Mineral 41 Determination of Trace Concentrations of Chlorate, Bromate, and Chlorite in Bottled Natural Mineral Waters

42 Table 2. Concentrations in mg/l of Cations and Anions in the Investigated Mineral Water Samples Mineral Na + K + Mg 2+ Ca 2+ F - C l- - NO 3 - HCO 3 2- SO 4 Water A a B a 5.0 < C 4.2 a < a Not specified Mineral water A B C Mineral water A B C Table 3. Recoveries of Disinfection Byproduct Anions in Natural Mineral Waters Using the IonPac AS19 Column Analyte Amount found (μg/l) Amount added (μg/l) Recovery (%) Chlorite <MDL Bromate <MDL Chlorate Chlorite <MDL Bromate <MDL Chlorate <MDL Chlorite <MDL Bromate <MDL Chlorate <MDL Table 4. Recoveries of Disinfection Byproduct Anions in Natural Mineral Waters Using the IonPac AS23 Column Analyte Amount found (μg/l) Amount added (μg/l) Recovery (%) Chlorite <MDL Bromate <MDL Chlorate Chlorite <MDL Bromate <MDL Chlorate <MDL Chlorite <MDL Bromate <MDL Chlorate <MDL operator data obtained using the IonPac AS19 and AS23 columns, respectively, for trace concentrations of DBP anions in three European natural mineral water samples. As shown, chlorite and bromate were notdetected in any of the samples analyzed,whereas only a trace concentration of chlorate was detected in mineral water A. To determine the accuracy of the method, the samples were spiked with 5 μg/l bromate and 10 μg/l each of chlorite and chlorate. Calculated recoveries for the spiked mineral water samples were in the content must be maintained at a constant level and no minerals may be added to the water. 18 In Europe, mineral water is defined as microbiologically wholesome water, originating from an underground water table or deposit and emerging from a spring tapped at one or more natural or bored exits. It can contain less than 50 ppm TDS. 19 The total mineral content of the waters can vary significantly, with higher mineral concentrations generally appearing in Russia, the Baltic States, and Germany. The differences between regions are most likely a result of differences in the overall compositions of the waters and the geological locations. 20 In this application, three natural mineral waters from different European countries with TDSs that varied significantly from 136 to 2359 ppm were evaluated. The properties of the investigated water samples are summarized in Table 2. As shown, the ionic strength of mineral water C is significantly higher than observed in typical drinking waters. The absence of bromate in the bottled mineral waters analyzed indicated that ozonation was not used for disinfection. Tables 3 and 4 summarize typical recoveries for singlerange of 86 97% and % using the IonPac AS19 and AS23 columns, respectively. The analyte recoveries using either a hydroxide or carbonate/bicarbonate eluent were within the acceptable range of % according to the criteria described in EPA Method Figure 3 compares chromatograms of mineral water A 42 Determination of Trace Concentrations of Chlorate, Bromate, and Chlorite in Bottled Natural Mineral Waters

43 Column: Eluent: A) IonPac AG19, AS19, 4 mm B) IonPac AG23, AS23, 4 mm A) 10 mm KOH 0-10 min, mm min, 45 mm min B) 4.5 mm potassium carbonate/ 0.8 mm potassium bicarbonate Eluent Source: A) EGC-KOH with CR-ATC B) EGC-K 2 CO 3 + EPM Temperature: 30 C Flow Rate: 1.0 ml/min Inj. Volume: 250 µl µs µs A B Detection: Suppressed conductivity ASRS ULTRA II, recycle mode A) Recycle mode B) External mode Peaks: Fluoride 2. Chloride 3. Nitrite 4. Chlorate A 4.4 B 4.6 µg/l 5. Bromide 6. Nitrate 7. Carbonate 8. Sulfate 9. Phosphate Figure 3. Comparison of the A) IonPac AS19 and B) IonPac AS23 columns for the separation of DPB anions in mineral water A. using the IonPac AS19 and AS23 columns. Figure 4 shows the same chromatograms spiked with 5 μg/l bromate and 10 μg/l each of chlorite and chlorate, which resulted in good recoveries for both eluents. Although bromide was not quantified in this study, the estimated concentrations were approximately 16 μg/l in mineral waters A and B and 2 μg/l in mineral water C. Therefore, ozonation of mineral waters A and B could potentially produce bromate. To demonstrate the applicability of detecting bromate at concentrations significantly less than the 3 μg/l European regulatory limit for ozonated mineral waters, mineral water A was spiked with 0.5 μg/l bromate (Figure 5). As shown, bromate can be observed easily at this concentration, with good peak-to-peak baseline noise of ns. 9 8 Column: Eluent: A) IonPac AG19, AS19, 4 mm B) IonPac AG23, AS23, 4 mm A) 10 mm KOH 0-10 min, mm min, 45 mm min B) 4.5 mm potassium carbonate/ 0.8 mm potassium bicarbonate Eluent Source: A) EGC-KOH with CR-ATC B) EGC-K 2 CO 3 + EPM Temperature: 30 C Flow Rate: 1.0 ml/min Inj. Volume: 250 µl µs µs A B Detection: Suppressed conductivity ASRS ULTRA II, recycle mode A) Recycle mode B) External mode Peaks: 1. Fluoride 2. Chlorite A 8.8 B 11.3 µg/l 3. Bromate Chloride Nitrite 6. Chlorate Bromide 8. Nitrate 9. Carbonate 10. Sulfate 11. Phosphate CONCLUSION The IonPac AS19 column using an electrolytically generated hydroxide eluent was compared to the AS23 column using an electrolytically generated carbonate/ bicarbonate eluent for the determination of trace concentrations of DBP anions in natural mineral waters. The improved sensitivity using a hydroxide eluent allowed the detection of lower concentrations of bromate, a potential human carcinogen, in drinking waters. Therefore, the IonPac AS19 with an electrolytically generated hydroxide eluent is recommended for laboratories that must comply with EU Directive 2003/40/EC, which permits a maximum of 3 μg/l bromate in mineral waters treated with ozone. The use of either the IonPac AS19 column with a hydroxide eluent or Figure 4. Comparison of the A) IonPac AS19 and B) IonPac AS23 columns for the separation of trace concentrations of common anions and DPB anions spiked in mineral water A. 43 Determination of Trace Concentrations of Chlorate, Bromate, and Chlorite in Bottled Natural Mineral Waters

44 Column: IonPac AG19, AS19, 4 mm Eluent: 10 mm KOH 0-10 min, mm min, 45 mm min Eluent Source: EGC-KOH with CR-ATC Temperature: 30 C Flow Rate: 1.0 ml/min Inj. Volume: 250 µl Detection: Suppressed conductivity ASRS ULTRA II, recycle mode 0.5 µs 1 4 Peaks: 1. Fluoride 2. Chlorite 1.0 µg/l 3. Bromate Chloride 5. Nitrite 6. Chlorate Bromide 8. Nitrate 9. Carbonate 10. Sulfate 11. Phosphate Figure 5. Chromatogram of mineral water A spiked with 1 µg/l each chlorite and chlorate and 0.5 µg/l bromate IonPac AS23 column with a carbonate/bicarbonate eluent provides the required sensitivity to meet the maximum permissible limit of 10 μg/l bromate currently required by most regulatory agencies. Both columns demonstrated good resolution between bromate and chloride and comparable recovery for mineral water samples spiked with known concentrations of chlorite, bromate, and chlorate. In addition, hydroxide or carbonate/bicarbonate eluents can be generated on-line from deionized water, freeing the operator from manually preparing eluents. This increases the automation, ease-of-use, and reproducibility between analysts and laboratories. REFERENCES 1. Rodwan, J.G. Reprint from The Bottled Water Reporter, Aug/May 2005, available at beveragemarketing.com/news3e.htm. 2. U.S. EPA, Water Health Series: Bottled Water Basics, U.S. Environmental Protection Agency, Cincinnati, OH, Sept U.S. EPA, Occurrence Assessment for the Final Stage 2 Disinfectants and Disinfection Byproducts Rule, Document No. 815-R , U.S. Environmental Protection Agency, Cincinnati, OH, December World Health Organization, International Programme on Chemical Safety-Environmental Health Criteria 216, Disinfectants and Disinfection By-Products, Geneva, Switzerland, Thompson, K.C.; Guinamant, J.L.; Elwaer, A.R.; McLeod, C.W.; Schmitz, F.; De Swaef, G.; Quevauviller, P. Interlaboratory Trial to Determine the Analytical State-of-the-Art of Bromate Determination in Drinking Water. J. Environ. Monit., 2000, 2, Wagner, H.P.; Pepich, B.V.; Hautman, D.P.; Munch, D.J. Analysis of 500 ng/l Levels of Bromate in Drinking Water by Direct Injection Suppressed Ion Chromatography Coupled with a Single, Pneumatically Delivered Post-Column Reagent. J. Chromatogr. A, 1999, 850, US EPA. National Primary Drinking Water Regulations: Disinfectants and Disinfection Byproducts. Fed. Regist. 1998, 63 (241), European Parliament and Council Directive No. 98/83/EC, Quality of Water Intended for Human Consumption, World Health Organization, Bromate in Drinking Water Background Document for the Development of WHO Guidelines for Drinking Water Quality, Food and Drug Administration. Beverages: Bottled Water. Fed. Reg. 2001, 66 (60), Determination of Trace Concentrations of Chlorate, Bromate, and Chlorite in Bottled Natural Mineral Waters

45 11. European Parliament and Council Directive No. 2003/40/EC, Establishing the List, Concentration Limits and Labeling Requirements for the Constituents of Natural Mineral Waters and the Conditions for Using Ozone-Enriched Air for the Treatment of Natural Mineral Waters and Spring Waters, Hautman, D. P.; Munch, D. J. Method 300.1: Determination of Inorganic Anions in Drinking Water by Ion Chromatography. U. S. Environmental Protection Agency, Cincinnati, OH, US EPA. National Primary Drinking Water Regulations: Stage 2 Disinfectants and Disinfection Byproducts Rule. Fed. Reg. 2006, 71 (2), Wagner, H. P.; Pepich, B. V.; Hautman, D. P.; Munch, D. J. Method 317.0, rev 2: Determination of Inorganic Oxyhalide Disinfection Byproducts in Drinking Water Using Ion Chromatography with the Addition of a Postcolumn Reagent for Trace Bromate Analysis. U. S. Environmental Protection Agency, Cincinnati, OH, Wagner, H. P.; Pepich, B. V.; Hautman, D. P.; Munch, D. J.; Salhi, E.; von Gunten, U. Method 326.0: Determination of Inorganic Oxyhalide Disinfection By-Products in Drinking Water Using Ion Chromatography Incorporating the Addition of a Suppressor Acidified Postcolumn Reagent for Trace Bromate Analysis. U. S. Environmental Protection Agency, Cincinnati, OH, Creed, J. T.; Brockhoff, C. A.; Martin. T. D. Method 321.8: Determination of Bromate in Drinking Waters by Ion Chromatography Inductively Coupled Plasma Mass Spectrometry. U.S. Environmental Protection Agency, Cincinnati, OH, Dionex Corporation. Determination of Trace Concentrations of Oxyhalides and Bromide in Municipal and Bottled Waters Using a Hydroxide- Selective Column with a Reagent-Free Ion Chromatography System. Application Note 167, LPN 1662; Sunnyvale, CA Dionex Corporation. Determination of Trace Concentrations of Disinfection By-Product Anions and Bromide in Drinking Water Using Reagent-Free Ion Chromatography Followed by Postcolumn Addition of o-dianisidine for Trace Bromate Analysis. Application Note 168, LPN 1706; Sunnyvale, CA Dionex Corporation. Determination of Disinfection By-Product Anions and Bromide in Drinking Water Using a Reagent- Free Ion Chromatography System Followed by Postcolumn Addition of an Acidified On-Line Generated Reagent for Trace Bromate Analysis. Application Note 171, LPN 1767; Sunnyvale, CA Posnick, L.M.; Kim, H. Bottled Water Regulation and the FDA. Food Safety Magazine, Aug/Sept 2002, 7(13 15). 21. European Parliament and Council Directive No. 80/777/EEC, The Approximation of the Laws of the Member States Relating to the Exploitation and Marketing of Natural Mineral Waters, Misund, A.; Frengstad, B.; Siewers, U.; Reimann, C. Variation of 66 Elements in European Bottled Mineral Waters. Sci. Total Environ. 1999, 243/244, Determination of Trace Concentrations of Chlorate, Bromate, and Chlorite in Bottled Natural Mineral Waters

46 Application Note 167 Determination of Trace Concentrations of Oxyhalides and Bromide in Municipal and Bottled Waters Using a Hydroxide-Selective Column with a Reagent-Free Ion Chromatography System INTRODUCTION All drinking water municipalities share the same goal of providing their communities with a reliable source of safe drinking water. To achieve this goal, most water systems must treat their water. The type of treatment used varies depending on the size, source, and water quality. 1 Disinfection protects public water systems (PWSs) from potentially dangerous microbes. The most common chemical disinfectants are chlorine, chlorine dioxide, chloramine, and ozone. 1,2 These chemical disinfectants can react with natural organic and inorganic matter in the source water to produce disinfection by-products (DBPs) that are potentially harmful to humans. For example, chlorination of drinking water can produce trihalomethanes, haloacetic acids, and chlorate. While chlorine dioxide treatment generates the inorganic oxyhalide DBPs chlorite and chlorate, and the presence of chloramine has also been known to generate chlorate. 2 Ozone reacts with naturally occurring bromide to produce bromate. The International Agency for Research on Cancer has identified bromate as an animal carcinogen and potential human carcinogen. 3 The World Health Organization (WHO) has estimated 4 an excess lifetime cancer risk of 10 5 for drinking water containing bromate at 3 µg/l. * * Probable increase in deaths due to a cancer, 10 5 = 1 in 100,000 people From July 1997 to December 1998, the U.S. Environmental Protection Agency (EPA) documented the occurrence of bromate and other DBPs through a comprehensive collection of sampling data mandated by the Information Collection Rule (ICR). 5 The ICR required that PWSs serving 100,000 or more connections report the concentration of target microorganisms present, the removal process used, and the concentration of DBPs present in their drinking water. In 1998, the EPA set the maximum contaminant level (MCL) for bromate at 10 µg/l and chlorite at 1000 µg/l under the Disinfectants/ Disinfection By-Products (D/DBP) Stage 1 Rule. 6 This rule resulted in the promulgation of EPA Method as an update to Method Method reduced the detection limit for bromate from 20 to 1.4 µg/l to allow the PWSs laboratories to meet the MCL requirement set by the EPA. 7 The European Union (EU Directive 98/83/EC) also proposed the same regulatory value of 10 µg/l bromate (previously at 50 µg/l) in drinking water. 8 The U.S. EPA reconvened in 2003 to establish the Stage 2 Rule of the D/DBP. Based on a thorough evaluation, the EPA could not estimate the additional benefits of reducing the MCL for bromate. Therefore, this rule resulted in no changes to the current MCL for either chlorite or bromate. However, additional methods for determining low µg/l bromate were promulgated 46 Determination of Trace Concentrations of Oxyhalides and Bromide in Municipal and Bottled Waters Using a Hydroxide-Selective Column with a Reagent-Free Ion Chromatography System

47 under the Stage 2 Rule and included ion chromatography (IC) with postcolumn reaction (EPA Methods and 326.0) and IC/ICP-MS (EPA Method 321.8). The addition of these methods resulted in improved sensitivity and selectivity for bromate. 9 Recently, the WHO reduced their bromate guideline value from 25 µg/l to a provisional value of 10 µg/l bromate. 4 This change resulted from the availability of improved analytical methods capable of determining low-µg/l concentrations of bromate in environmental waters. Unlike tap water, bottled water is treated as a food product in the U.S. and therefore regulated by the U.S. Food and Drug Administration (FDA). Bottled water is an increasingly popular product in the U.S. From 1997 to 2002, bottled water sales increased from roughly 6% to 13% per year of total beverage sales, according to the Beverage Marketing Corporation. 10 Because some bottled water companies use ozone or other disinfection treatments, the FDA adopted the EPA s MCLs for chlorite and bromate and the analytical methods used to monitor these contaminants in public drinking water. 11 The FDA also requires that bottled water manufacturers monitor their finished product for these contaminants at least once each year under current good manufacturing practice as stated in part 129 of the Code of Federal Regulations (21 CFR part 129). Previous methods developed for determining low- µg/l concentrations of bromate by direct injection have focused primarily on using columns specifically designed for carbonate eluents. 12,13 Columns designed for use with hydroxide eluents have not been widely used for the determination of trace bromate in environmental waters due to their lack of appropriate column selectivity and the difficulty in preparing contaminant-free hydroxide eluents. The introduction of electrolytic eluent generation has not only eliminated the difficulty in preparing hydroxide eluents, but has simplified the development of optimized methods. In this application note, we use the IonPac AS19, a column specifically designed for use with hydroxide eluents and developed with an optimized selectivity for the determination of trace DBPs and bromide in environmental waters. We describe the linearity, method detection limits, and the recovery and precision of spiked municipal and bottled waters. EQUIPMENT A Dionex ICS-2000 Reagent-Free Ion Chromatography (RFIC ) * System was used in this work. The ICS-2000 is an integrated ion chromatograph and consists of: Eluent Generator Column Heater Pump with Degasser EluGen EGC II KOH Cartridge (Dionex P/N ) CR-ATC (Dionex P/N ) AS50 Autosampler Chromeleon Chromatography Workstation *This application note is also applicable to other RFIC systems. REAGENTS AND STANDARDS Deionized water, Type I reagent-grade, 18 MΩ-cm resistivity or better Sodium and Potassium salts, ACS reagent-grade or better, for preparing anion standards (VWR or other) Fluoride standard 1000 mg/l, 100 ml (Dionex P/N ) Chloride standard 1000 mg/l, 100 ml (Dionex P/N ) Sulfate standard 1000 mg/l, 100 ml (Dionex P/N ) Bromide standard 1000 mg/l, 100 ml (Ultra Scientific, VWR P/N ICC-001) Sodium Chlorite, 80% (Fluka Chemical Co.) Sodium Bromate (EM Science, VWR P/N EM SX0385-1) Ethylenediamine, 99% (Sigma-Aldrich) CONDITIONS Columns: Eluent: IonPac AS19 Analytical, mm (Dionex P/N ) IonPac AG19 Guard, 4 50 mm (Dionex P/N ) 10 mm KOH from 0 to 10 min, mm from 10 to 25 min * Eluent Source: ICS-2000 EG with CR-ATC Flow Rate: 1.0 ml/min Temperature: 30 C Injection: 250 µl 47 Determination of Trace Concentrations of Oxyhalides and Bromide in Municipal and Bottled Waters Using a Hydroxide-Selective Column with a Reagent-Free Ion Chromatography System

48 Detection: Suppressed conductivity, ASRS Background Conductance: <1 µs System Backpressure: Run Time: ULTRA II, 4 mm (Dionex P/N ) AutoSuppression Recycle Mode 130 ma current ~2200 psi 30 min * Method returns to 10 mm KOH for 3 min prior to injection PREPARATION OF SOLUTIONS AND REAGENTS Stock Standard Solutions For several of the anions of interest, 1000-mg/L standard solutions can be purchased from Dionex or other commercial sources. When commercial standards are not available, 1000-mg/L standards can be prepared by dissolving the appropriate amounts of the required analytes in 100 ml of deionized water according to Table 1. Stock standards for most anions are stable for at least 6 months when stored at 4 C. The chlorite standard is only stable for two weeks when stored protected from light at 4 C. The nitrite and phosphate standards are only stable for one month when stored at 4 C. Working Standard Solutions Dilute working standard solutions were prepared using the 1000-mg/L stock standards. Working standards containing less than 100 µg/l anions should be prepared fresh daily. Seven levels of calibration standards were used in this study for chlorite, chlorate, and bromide to cover the expected concentration range found in typical environmental samples. The bromate calibration curve was prepared using eight calibration standards. Additional anions listed in Table 1 were used to prepare a simulated drinking water sample containing 1 ppm fluoride, 50 ppm chloride, 0.1 ppm nitrite, 10 ppm nitrate, 100 ppm carbonate, 50 ppm sulfate, and 0.1 ppm phosphate. Preservation Solution Dilute 2.8 ml of 99% ethylenediamine (EDA) to 25 ml with deionized water according to section 7.4 in EPA Method to prepare a 100-mg/mL solution of EDA. Use 50 µl of this solution per 100 ml of standard or sample so that the final concentration is 50 mg/l. Table 1. Masses of Compounds Used to Prepare 100 ml of 1000-mg/L Ion Standards Analyte Compound Amount (g) Fluoride Sodium fluoride (NaF) Chlorite Sodium chlorite (NaClO 2 ), 80% Bromate Sodium bromate (NaBrO 3 ) Chloride Sodium chloride (NaCI) Nitrite Sodium nitrite (NaNO 2 ) Chlorate Sodium chlorate (NaCIO 3 ) Bromide Sodium bromide (NaBr) Nitrate Sodium nitrate (NaNO 3 ) Sulfate Sodium sulfate (Na 2 SO 4 ) Phosphate Potassium phosphate, monobasic (KH 2 PO 4 ) Sample Preparation Filter samples, as necessary, through a 0.45-µm syringe filter, discarding the first 300 µl of the effluent. To prevent degradation of chlorite or the formation of bromate from hypobromous acid/hypobromite, preserve the samples by adding 50 µl of EDA preservation solution per 100 ml of sample. RESULTS AND DISCUSSION EPA Method Part B currently specifies an IonPac AS9-HC column using a carbonate eluent and suppressed conductivity detection for the determination of trace DBP anions and bromide in environmental waters, such as drinking water, surface water, and groundwater. 7 The use of the IonPac AS9-HC column in EPA Method (B) significantly improved the determination for trace bromate compared to the AS9-SC specified in Method 300.0, Part B. 14 The AS9-HC allowed for detection limits to 1.4 µg/l bromate with a 200-µL injection volume, even in the presence of excess chloride. However, the use of a hydroxide eluent for the determination of trace bromate is more appealing than carbonate eluents. Hydroxide eluent has significantly lower suppressed background conductivity, lower noise, and therefore lower detection limits compared to carbonate eluents. Previously, we described the advantages of hydroxide over carbonate eluents for the determination of common anions. 15 Therefore, similar advantages for bromate should be expected using a column with an appropriate selectivity combined with a hydroxide eluent. 48 Determination of Trace Concentrations of Oxyhalides and Bromide in Municipal and Bottled Waters Using a Hydroxide-Selective Column with a Reagent-Free Ion Chromatography System

49 The IonPac AS19 is a high-capacity, hydroxideselective column specifically designed for the determination of trace bromate and other oxyhalides using a large-volume injection. The novel polymer chemistry of the AS19 yields a higher capacity of 240 µequiv/ column compared to the AS9-HC (190 µeq/column). The AS19 stationary phase is based on a new hyperbranched anion-exchange condensation polymer that is electrostatically attached to the surface of a wide-pore polymeric substrate. The AS19 selectivity and capacity are optimized to achieve good resolution between bromate and chloride. Unlike previous IonPac columns, the anion-exchange resin of the AS19 contains alternating treatments of epoxy and amines to produce a coating that grows directly off the surface-sulfonated substrate. The number of alternating coating cycles results in a carefully controlled ion-exchange capacity with an extremely hydrophilic polymer. Therefore, the column has excellent selectivity for hydroxide eluents, allowing lower concentrations of hydroxide to be used. 16 Figure 1 shows a separation of common anions and disinfection by-product anions separated within 30 min using the AS19 column with a hydroxide gradient. As this figure shows, the AS19 achieves excellent resolution between bromate and chloride, making it ideal for determining low concentrations of bromate in municipal and bottled water samples. Linearity and Method Detection Limits Before conducting any sample analyses, the linear calibration range, MDLs, and acceptable performance of a quality control sample (QCS) should be demonstrated. Initially, a seven-point calibration range was used for chlorite, chlorate, and bromide, whereas eight calibration points were used for bromate. MDLs for each anion listed in EPA Method 300.1, Part B were determined by performing seven replicate injections of reagent water fortified at a concentration of three to five times the estimated instrument detection limits. In addition, the MDLs were also determined by fortifying the same concentration of anions in a simulated drinking water sample. Table 2 shows typical calculated MDLs in reagent water and simulated drinking water using the IonPac AS19 column combined with an electrolytic eluent generator and a 250-µL injection. In comparing the detection limits in the two matrices, the results showed 60.0 µs 0.5 Column: IonPac AG19 and AS19, 4 mm Eluent: 10 mm KOH 0 10 min, mm min Eluent Source: ICS-2000 EG with CR-ATC Flow Rate: 1 ml/min Temperature: 30 C Inj. Volume: 25 µl Detection: ASRS ULTRA II, 4 mm recycle mode, 130 ma Peaks: 1. Fluoride 3 mg/l 2. Chlorite Bromate Chloride 5 5. Nitrite Chlorate Bromide Nitrate Carbonate 10. Sulfate Phosphate Figure 1. Separation of common anions and disinfection by-product anions on the IonPac AS19 column. 5 no significant difference. The only exception was the calculated MDL for bromate in simulated drinking water was only slightly greater, as expected, because increasing concentrations of chloride will affect the determination of low concentrations of bromate. The calculated MDL for bromate using this method was 0.34 µg/l, approximately 70% lower than previously reported with the AS9-HC column at comparable injection volumes. 12 The lower detection limit results from the excellent peak efficiencies Table 2. Method Detection Limits for Oxyhalides and Bromide in Reagent Water and Simulated Drinking Water Using the IonPac AS19 Columna Analyte MDL Standard Calculated MDL b in Reagent Water Calculated MDL b in Simulated Drinking Water Chlorite Bromate Chlorate Bromide a 250-µL injection volume b MDL = st s,99 where t s,99 = 3.14 for n = 7 49 Determination of Trace Concentrations of Oxyhalides and Bromide in Municipal and Bottled Waters Using a Hydroxide-Selective Column with a Reagent-Free Ion Chromatography System

50 of the AS19 combined with low noise and exceptionally low suppressed background conductivities obtained by using an electrolytically generated hydroxide eluent. These results demonstrate the significant advantages of using an RFIC system for the determination of trace bromate. Figure 2 shows a separation of an MDL standard prepared in reagent water. As shown, bromate concentrations as low as 1.5 µg/l are easily detected by this method. Table 3 shows the linear concentration ranges investigated, the coefficients of determination (r 2 ), and the retention time and peak area precisions of a QCS based on 10 replicate injections. The excellent retention time stability and peak area precisions are consistent with results typically encountered when using an electrolytically generated high-purity potassium hydroxide eluent. The data presented in Table 3 demonstrate the advantages of using a hydroxide-selective column for routine applications, such as the determination of oxyhalides and bromide in environmental waters. The advantages of using IC with a hydroxide eluent are improved linearity, lower background conductivity, and improved method detection limits when compared with conventional IC columns that use carbonate eluents, such as the IonPac AS9-HC. The use of an electrolytically generated potassium hydroxide eluent further simplifies the method by eliminating the time required to manually prepare eluents and by reducing the time required for method development. Effect of Column Overloading The effect of sample overload on the IonPac AS19 column was evaluated as part of this study. One of the many challenges encountered when determining trace concentrations of bromate is the potential presence of a high sample chloride concentration. In addition to chloride, a high concentration of other anions can together reduce the amount of bromate recovered from a sample. For most environmental samples, chloride, sulfate, and carbonate are generally present at the greatest amounts with respect to other common anions. For this study, we chose a 250-µL sample injection for the analyses because this volume provided us the sensitivity necessary to achieve low-ppb detection of bromate and reduced the likelihood of overloading the column when analyzing high-ionic-strength samples. 0.5 µs 0.0 Column: IonPac AG19 and AS19, 4 mm Eluent: 10 mm KOH 0 10 min, mm min Eluent Source: ICS-2000 EG with CR-ATC Flow Rate: 1 ml/min Temperature: 30 C Inj. Volume: 250 µl Detection: ASRS ULTRA II, 4 mm recycle mode, 130 ma 0.05 µs Peaks: 1. Chlorite 1.0 µg/l 2. Bromate Chlorate Bromide Figure 2. Separation of DBP anions and bromide method detection limit standard. Table 3. Linearity and Retention Time and Peak Area Precisions Obtained Using the IonPac AS19 Column a Analyte Range Linearity (r 2 ) Retention Time Precision (%RSD b ) Peak Area Precision (%RSD) Chlorite < Bromate < Chlorate < Bromide < a Dionex ICS-2000 Reagent-Free IC system with a 250-mL injection volume b RSD = relative standard deviation, n = Determination of Trace Concentrations of Oxyhalides and Bromide in Municipal and Bottled Waters Using a Hydroxide-Selective Column with a Reagent-Free Ion Chromatography System

51 To determine the effect of chloride on bromate recovery, a series of increasing concentrations of chloride was added to Sunnyvale drinking water. Figure 3 illustrates the effect of increasing concentrations of chloride on the recovery of 5 µg/l bromate. As shown, the recovery of bromate is acceptable in the presence of ~150 ppm chloride. Above this concentration, the bromate significantly decreases to an unacceptable recovery (e.g., <75%). Based on this analysis, the IonPac AS19 can tolerate up to ~150 ppm chloride, resulting in a bromateto-chloride ratio of 1:30,000, comparable to the AS9- HC column. 12 A similar experiment was performed by increasing the sulfate concentration without any additional chloride added. This experiment demonstrated very little change in the bromate recovery for up to 200 ppm sulfate (results not shown). However, high concentrations of chloride and sulfate can combine to have a greater impact on reducing the bromate recovery. Equal concentrations of chloride and sulfate (up to 120 ppm each) were added to Sunnyvale drinking water, resulting in a 75% bromate recovery. However, most drinking water samples contain significantly less chloride and sulfate than the concentrations included in this study. For example, 18 of the samples examined contained chloride concentrations ranging from <0.1 to 70 ppm and sulfate from <0.1 to 60 ppm. Therefore, almost all samples can be easily analyzed using a 250-µL injection volume, while the column can tolerate 500-µL injections of low- to moderate-ionicstrength samples. Figure 4 shows a chromatogram of a 500-µL injection of Sunnyvale drinking water spiked with oxyhalides and bromide. As shown, bromate was well resolved from chloride with bromate recovered at nearly 100%. Figure 5 compares 250- to 500-µL injection volumes for a simulated drinking water sample. The 500-µL injection volume caused some column overloading and therefore a lower bromate recovery of ~74%. However, a 250-µL injection of the same sample significantly improved the recovery of bromate to 92%. Therefore, a 250-µL injection is recommended for most sample analyses. The effect of column overloading is most prevalent on early-eluting peaks, observed by increased peak broadening and lower recoveries, as demonstrated in this example. Bromate Recovery (%) µs Chloride conc. (ppm) Figure 3. Effect of increasing the chloride concentration on the recovery of 5 µg/l bromide. Column: IonPac AG19 and AS19, 4 mm Eluent: 10 mm KOH 0 10 min, mm min Eluent Source: ICS-2000 EG with CR-ATC Flow Rate: 1 ml/min Temperature: 30 C Inj. Volume: 500 µl Detection: ASRS ULTRA II, 4 mm recycle mode, 130 ma Figure 4. Determination of DBP anions and bromide spiked in drinking water A using a 500-µL injection volume. 3 4 Peaks: 1. Chlorite 23 µg/l 2. Bromate 5 3. Chlorate Bromide Determination of Trace Concentrations of Oxyhalides and Bromide in Municipal and Bottled Waters Using a Hydroxide-Selective Column with a Reagent-Free Ion Chromatography System

52 Column: Eluent: Eluent Source: Flow Rate: Temperature: 30 C Inj. Volume: (A) 250 µl (B) 500 µl Detection: 1.50 µs µs 0.10 A B IonPac AG19 and AS19, 4 mm 10 mm KOH 0 10 min, mm min ICS-2000 EG with CR-ATC 1 ml/min ASRS ULTRA II, 4 mm recycle mode, 130 ma Peaks: 1. Fluoride 1.0 mg/l 2. Chlorite Bromate Chloride Nitrite Chlorate Bromide Nitrate Carbonate Sulfate Phosphate , 10 9, Accuracy and Precision The performance of the IonPac AS19 was also evaluated through a single-operator precision and bias study using spiked municipal and bottled water samples. Table 4 shows typical recoveries for single-operator data obtained using the IonPac AS19 column for trace concentrations of DBPs and bromide in environmental waters. Most anions demonstrated acceptable recoveries (i.e., %) according to the criteria outlined in EPA Method However, drinking water E resulted in an exceptionally lower recovery for bromide, regardless of the amount of bromide spiked in the sample. Section of EPA Method states, If the recovery of any analyte falls outside the LFM [Laboratory Fortified Matrix] recovery range and the laboratory performance for that analyte is shown to be in control, the recovery problem encountered with the LFM is judged to be either matrix or solution related, not system related. Therefore, the sample was labeled as suspect/matrix to indicate that the poor recovery of bromide was sample related and not system related. Figure 5. Comparison of simulated drinking water using (A) 250-µL injection and (B) 500-µL injection. Analyte a Suspect/matrix b Sample diluted 1:1 c Calculated amounts Table 4. Recoveries of Trace Oxyhalides and Bromide Spiked into Environmental Waters Drinking Water A Drinking Water B Drinking Water C Drinking Water D Amount Found Amount Added Recovery (%) Amount Found Amount Added Recovery (%) Amount Found Amount Added Recovery (%) Amount Found Amount Added Chlorite <MDL <MDL Bromate <MDL <MDL <MDL Chlorate Bromide Drinking Water E Surface Water Shallow Well Water b Well Water b Analyte Amount Found Amount Added Recovery (%) Amount Found Amount Added Recovery (%) Amount Found Amount Added Recovery (%) Amount Found c Amount Added Chlorite <MDL <MDL <MDL Bromate <MDL <MDL <MDL Chlorate <MDL <MDL Bromide <MDL a <MDL Recovery (%) Recovery (%) 52 Determination of Trace Concentrations of Oxyhalides and Bromide in Municipal and Bottled Waters Using a Hydroxide-Selective Column with a Reagent-Free Ion Chromatography System

53 Column: IonPac AG19 and AS19, 4 mm Peaks: (A) 2. Bromate Eluent: 10 mm KOH 0 10 min, mm min Eluent Source: ICS-2000 EG with CR-ATC Flow Rate: 1 ml/min (B) 1. Chlorite Temperature: 30 C Inj. Volume: 250 µl Detection: ASRS ULTRA II, 4 mm recycle mode, 130 ma Sample Prep: 1:1 dilution 1.50 A µs 4 3. Chlorate Bromide Bromate Chlorate Bromide µg/l 18 µg/l Column: IonPac AG19 and AS19, 4 mm Eluent: 10 mm KOH 0 10 min, mm min Eluent Source: ICS-2000 EG with CR-ATC Flow Rate: 1 ml/min Temperature: 30 C Inj. Volume: 250 µl Detection: ASRS ULTRA II, 4 mm recycle mode, 130 ma 1.50 µs A 3 4 Peaks: (A) 3. Chlorate 118 µg/l 4. Bromide 200 (B) 1. Chlorite 19 µg/l 2. Bromate 5 3. Chlorate Bromide B B 4 µs µs Figure 6. Determination of DBP anions and bromide in (A) shallow well water and (B) spiked shallow well water using the IonPac AS19 column. Figure 7. Determination of DBP anions and bromide in (A) drinking water B and (B) spiked drinking water B using the IonPac AS19 column. Due to the high ionic strength of the well water samples, both were diluted 1:1 to avoid column overloading. The estimated chloride and sulfate concentrations were 160 and 270 ppm, respectively, for the shallow well water, and 150 and 170 ppm, respectively, for the well water prior to dilution. These concentrations exceed the limits determined for this column during the sample overload study. Section in EPA Method states that sample dilution will alter your Minimum Reporting Limit (MRL) by a proportion equivalent to that of the dilution. In this study, dilution of the well water samples increased the bromate MRL from 1 to 2 µg/l. However, the adjusted MRL was still sufficient to report the 8 µg/l bromate detected in the diluted sample. Because this well water sample is not known to be treated, the presence of bromate was unexpected. The detection of bromate in the well water may result from contamination by a nearby site that originally contained a high concentration of the anion. Figure 6A shows a chromatogram of diluted shallow well water. Figure 6B shows the same well water sample spiked with µg/l of DBP anions and 200 µg/l of bromide. As shown, bromate was well resolved from the high concentration of chloride, resulting in a recovery of 101.1%. Figure 7 shows chromatograms of an unspiked and spiked drinking water B. This sample also demonstrates the excellent resolution and accuracy of analysis for the determination of trace DBP anions and bromide using an RFIC system. The calculated recoveries of the target analytes ranged from 96 to 106%. This study also included the analysis of a variety of bottled water samples randomly obtained from a local supermarket. A previous study conducted from in Canada found many bottled waters contained bromate, some at concentrations greater than 25 µg/l. 17 These results in combination with the increasing popularity of bottled water, led us to examine the presence of bromate in several different brands of bottled waters. More than half of the bottled waters tested in this study reported using ozonation as a form of treatment according to the bottle s label or company s web site (Table 5). 53 Determination of Trace Concentrations of Oxyhalides and Bromide in Municipal and Bottled Waters Using a Hydroxide-Selective Column with a Reagent-Free Ion Chromatography System

54 Table 5. Treatments Used for Different Bottled Waters Bottled Water Treatment 1 Natural spring water (no treatment) 2 UV light, RO a, ozonation 3 Ozonation 4 Natural mineral water (no treatment) 5 RO 6 Microfiltration, UV light, ozonation 7 Filtration 8 Microfiltration, ozonation 9 Natural spring water (no treatment) 10 Microfiltration, ozonation 11 Microfiltration, RO, DI b, ozonation 12 Ozonation a RO = reverse osmosis b DI = deionization Table 6 shows the amount found and the recoveries obtained using the AS19 column for trace concentrations of DBP anions and bromide spiked in the bottled waters. All target analytes demonstrated acceptable recoveries according to U.S. EPA Method Only four bottles tested contained some amount of bromate, with two of these near or slightly above the bromate MCL of 10 µg/l. No correlation was observed between the concentrations of bromide in the samples versus the amount of bromate detected. For example, bottled water #10 contained approximately 4 µg/l bromate, but no bromide was detected in the sample. However, the conversion of bromide to bromate upon ozonation is affected by several factors, such as the presence of natural organic matter, ph, temperature, and other variables. 2 As expected, most bottled waters analyzed contained appreciably less chloride and sulfate than tap water with estimated maximum concentrations of 8 and 30 ppm, respectively. Analyte Table 6. Recoveries of Trace Oxyhalides and Bromide Spiked into Bottled Waters Bottled Water 1 Bottled Water 2 Bottled Water 3 Bottled Water 4 Amount Found Amount Added Recovery (%) Amount Found Amount Added Recovery (%) Amount Found Amount Added Recovery (%) Amount Found Amount Added Recovery (%) Chlorite <MDL <MDL <MDL <MDL Bromate <MDL <MDL <MDL Chlorate <MDL <MDL Bromide <MDL Bottled Water 5 Bottled Water 6 Bottled Water 7 Bottled Water 8 Analyte Amount Found Amount Added Recovery (%) Amount Found Amount Added Recovery (%) Amount Found Amount Added Recovery (%) Amount Found Amount Added Recovery (%) Chlorite <MDL <MDL <MDL <MDL Bromate <MDL <MDL <MDL Chlorate <MDL <MDL Bromide Bottled Water 9 Bottled Water 10 Bottled Water 11 Bottled Water 12 Analyte Amount Found Amount Added Recovery (%) Amount Found Amount Added Recovery (%) Amount Found Amount Added Recovery (%) Amount Found Amount Added Recovery (%) Chlorite <MDL <MDL <MDL <MDL Bromate <MDL <MDL Chlorate <MDL <MDL <MDL Bromide <MDL <MDL <MDL Determination of Trace Concentrations of Oxyhalides and Bromide in Municipal and Bottled Waters Using a Hydroxide-Selective Column with a Reagent-Free Ion Chromatography System

55 The low ionic content of most bottled waters allows the use of larger injection volumes (500 µl or more). Figure 8 shows a 250-µL injection of an unspiked and spiked bottled water sample. The disinfection treatment used for this bottled water was UV radiation and ozonation. An unusually high amount of chlorate was detected in the sample, indicating that some form of chlorination may also be used for treatment. Bromate was detected at a slightly lower concentration than the EPA s MCL, possibly indicative of elevated levels of bromide in the source water. The recoveries for oxyhalide DBPs and bromide spiked in the sample ranged from ~97 to 107%, well within EPA Method specifications. The precision of the method using the AS19 column in combination with an electrolytic eluent generation was determined by performing 10 replicate injections of randomly selected samples spiked with trace concentrations of DBPs and bromide. Overall, the calculated peak area precisions varied from 0.21 to 1.78% with retention time precisions <0.04% for most target analytes. For bromate, the worst peak area precision observed was 1.78%. This number represents a deviation of only ±0.09 µg/l based on a sample containing 5 µg/l bromate. The high precision of this method is consistent with results typically found with an RFIC system. CONCLUSION IC with a hydroxide-selective IonPac AS19 column and an electrolytic eluent generator is an improved approach for determining trace concentrations of DBP anions and bromide in municipal and bottled water samples. The high-capacity AS19 column can be used with large-volume injections to detect lowppb concentrations of bromate, a potential human carcinogen, in many municipal and bottled waters. In addition, electrolytic generation of an ultrapure potassium hydroxide eluent, combined with the AS19 column, Column: IonPac AG19 and AS19, 4 mm Eluent: 10 mm KOH 0 10 min, mm min Eluent Source: ICS-2000 EG with CR-ATC Flow Rate: 1 ml/min Temperature: 30 C Inj. Volume: 250 µl Detection: ASRS ULTRA II, 4 mm recycle mode, 130 ma 1.50 µs µs 0.10 A B Peaks: (A) 2. Bromate 9.2 µg/l 3. Chlorate Bromide 2.5 (B) 1. Chlorite 20.0 µg/l 2. Bromate Chlorate Bromide Figure 8. Determination of DBP anions and bromide in (A) bottled water 6 and (B) spiked bottled water 6 using the IonPac AS19 column improves linearity, MDLs, precision, and resolution between bromate and chloride compared to the AS9-HC column described in EPA Method This approach also eliminates the need to manually prepare eluents and thereby increases the automation, ease of use, and reproducibility between analysts and laboratories. The U.S. EPA, Office of Water, has determined that the use of hydroxide eluents in EPA Method is acceptable for compliance monitoring under the Clean Water Act and Safe Drinking Water Act Determination of Trace Concentrations of Oxyhalides and Bromide in Municipal and Bottled Waters Using a Hydroxide-Selective Column with a Reagent-Free Ion Chromatography System

56 REFERENCES 1. Drinking Water Treatment; EPA 810-F ; U.S. Environmental Protection Agency, World Health Organization. Disinfectants and Disinfection By-Products; International Programme on Chemical Safety-Environmental Health Criteria 216; Geneva, Switzerland, Wagner, H. P.; Pepich, B. V.; Hautman, D. P.; Munch, D. J. J. Chromatogr., A 1999, 850, World Health Organization. Draft Guideline for Drinking Water Quality; Third ed., Fed. Regist. 1996, 61 (94), Fed. Regist. 1998, 63 (241), U.S. EPA Method 300.1; U.S. Environmental Protection Agency; Cincinnati, OH, European Parliament and Council Directive No. 98/83/EC, Quality of Water Intended for Human Consumption, Fed. Regist. 2003, 68 (159), Posnick, L. M.; Henry, K. Food Safety Magazine, Aug/Sept Fed. Regist. 2001, 66 (60), Jackson, L. K.; Joyce, R. J.; Laikhtman, M.; Jackson, P. E. J. Chromatogr., A 1998, 829, Dionex Corporation. Application Note 81 (LPN ); Sunnyvale, CA. 14. U.S. EPA Method 300.0; U.S. Environmental Protection Agency; Cincinnati, OH, Dionex Corporation. Application Note 154 (LPN 1539); Sunnyvale, CA. 16. Dionex Corporation. IonPac AS19 Anion-Exchange Column (data sheet) (LPN 1616); Sunnyvale, CA. 17. Lo, B.; Williams, D. T.; Subramanian, K. S. Am Lab. Feb. 1999, Letter to Dionex Corporation. U.S. Environmental Protection Agency, Office of Water; November 19, Determination of Trace Concentrations of Oxyhalides and Bromide in Municipal and Bottled Waters Using a Hydroxide-Selective Column with a Reagent-Free Ion Chromatography System

57 Application Note 196 Determination of Polycyclic Aromatic Hydrocarbons (PAHs) in Edible Oils by Donor-Acceptor Complex Chromatography (DACC)-HPLC with Fluorescence Detection Introduction Numerous polycyclic aromatic hydrocarbons (PAHs) are carcinogenic, making their presence in foods and the environment a health concern. Regulations around the world limit levels of a variety of PAHs in drinking water, food additives, cosmetics, workplaces, and factory emissions. PAHs also occur in charbroiled and dried foods, and may form in edible oils by pyrolytic processes, such as incomplete combustion of organic substances. PAHs in foods can also result from petrogenic contamination. The European Commission regulates the amounts of PAHs in foods, and has imposed a limit of 2.0 µg/kg for benzo[a]pyrene (BaP) in edible oils, as BaP was determined to be a good indicator of PAH contamination. 1 PAHs have traditionally been separated using HPLC and determined using UV, 2 fluorescence, 3,4 electrochemical, 5 and mass spectrometry (using atmospheric-pressure photoionization) 6 detection methods. After an oxygenation reaction, PAHs can also be determined by LC-MS/MS. 7 These methods of determining PAHs in edible oils require multiple manual sample preparation steps. One study of PAHs in over a dozen edible oils used a DMSO extraction followed by three extractions with cyclohexanone and cleanup with a silica column. 8 Another study of six edible oils used solid-phase extraction, but required solvent extraction steps before SPE and evaporation afterward. 9 These manual steps consume solvent, resources, and time. In recent years, donor-acceptor complex chromatography (DACC) has gained popularity for PAH analysis DACC stationary phases can be used for solid phase extraction (SPE), retaining PAHs while matrix components are flushed to waste. After elution of the analytes, solvent exchange is used to prepare the sample for HPLC analysis. Compared to traditional methods, this cleanup technique uses less solvent, is less labor intensive, and saves considerable time. 11 However, this approach still involves several manual samplehandling steps, and therefore still requires labor and is prone to errors. In 1996, Van Stijn et al. 12 developed an automated process for oil sample preparation and analysis. The setup consists of coupling a DACC cleanup column with an HPLC analytical column. This solution does not require manual cleanup and solves the previously described challenges. However, adopting the method for routine operation is difficult and requires advanced technical know-how to optimize the system configuration. This optimization can be time consuming. Furthermore, the described solution uses the autosampler software for system control and different software for data collection, instead of using an integrated chromatography data system for system control and monitoring. This leaves room for improvement in ease of operation, process monitoring and documentation, validation, reporting, and automated diagnosis. 57 Determination of Polycyclic Aromatic Hydrocarbons (PAHs) in Edible Oils by Donor-Acceptor Complex Chromatography (DACC)-HPLC with Fluorescence Detection

58 Maio et al. 13 adapted Van Stijn's solution to create a method for automated on-line determination of PAHs in edible oils that addresses the remaining challenges. This solution was performed on an HPLC system equipped with a dual-gradient HPLC pump and two switching valves, allowing on-line sample enrichment on a DACC column with HPLC analysis. On-line coupling of sample preparation and analysis eliminates the complex manual pretreatment required by traditional methods. This automation reduces unintentional errors and increases reproducibility. The analysis time per sample is approximately 80 min with the dual-gradient HPLC system, compared to 8 10 h with traditional methods. Moreover, this automated system can run 24 h a day, significantly increasing sample throughput and making this complex analysis routine. This application note describes the setup and method for determining PAHs in edible oils on-line using a Dionex UltiMate Dual-Gradient HPLC platform, based on the solution described by Maio, 13 and evaluates the method performance (i.e. linearities, detection limits, reproducibilities, recoveries, and carry over of autosampler). Equipment UltiMate Dual system consisting of: DPG-3600A pump with SRD-3600 Air Solvent Rack WPS-3000TSL autosampler TCC-3200 thermostatted column compartment with two 2p-6p valves RF2000 fluorescence detector Chromeleon 6.80 SP1 Chromatography Workstation Device configurations for the online DACC cleanup with analytical HPLC are as shown in Figures 1 to 3. In these figures, the upper valve is the right valve and lower valve is the left valve in the TCC Reagent and Standards Deionized water from a Milli-Q Gradient A10 Acetonitrile (CH 3 CN), HPLC grade (Fisher Scientific) Isopropanol, HPLC grade, (Fisher Scientific) Charcoal, activated granular (activated carbon), chemical pure grade (Shanghai Chemical Reagent Company) Mix of PAHs, EPA Sample for Method 610, 200 µg/ ml for each component, including phenanthrene, anthracene, fluoranthene, pyrene, benzo[a] anthracene, chrysene, benzo[b]fluoranthene, benzo[k]fluoranthene, benzo[a]pyrene, dibenzo[a,h] anthracene, benzo[g,h,i]perylene and indeno[1,2,3- cd]pyrene (Restek) Benzo[b]chrysene, 50 µg/ml, used as an internal standard (I.S.) (AccuStandard) Samples Two brands of olive oil (olive oil 1 and 2 from Italy and Spain, respectively), and one brand of sesame oil (from China) Conditions Analytical Columns: On-Line SPE Column: Mobile Phases: Two SUPELCOSIL LC-PAH columns, mm (Supelco Cat.# 58229) ChromSpher Pi, mm (Varian P/N: CP28159) A. Water B. Acetonitrile for both loading and analysis pumps C. Isopropanol for loading pump 1 ml/min 80 μl (100 μl injection loop) Flow Rate: Injection Volume: Column Temperature: 30 C Autosampler Temperature: 40 C Detection: Fluorescence (Table 4) 58 Determination of Polycyclic Aromatic Hydrocarbons (PAHs) in Edible Oils by Donor-Acceptor Complex Chromatography (DACC)-HPLC with Fluorescence Detection

59 Table 1 shows the gradient for on-line solid-phase extraction (SPE) using the loading pump, and Table 2 shows the gradient for separation using the analysis pump. Table 3 shows the valve switching timing. Because the maximum fluorescent responses of PAHs occur at different emission wavelengths, it is necessary to change the excitation and emission wavelengths based on individual PAH retention times. Table 4 shows the program for wavelength changes. Preparation of Standards and samples Purification of Olive Oil Used as Blank and Matrix Add 1 g activated carbon to 20 g olive oil, heat for 2 h at 60 C while stirring, then filter through a pleated filter. Pass filtrate through a membrane filter (0.45 µm, PTFE, Millex -LCR, Millipore) and store the resulting purified oil sample at 4 C. Preparation of Olive Oil Containing I.S. Used as Matrix To prepare a 0.25 µg/ml stock I.S. solution, add 995 µl isopropanol using a 1-mL pipette to a 2-mL vial, and add 5 µl of 50 µg/ml I.S. oil using a 10-µL syringe. Add 40 µl of the 0.25 µg/ml stock I.S. solution, using a 10-µL syringe, to about 10 g of the purified olive oil used as blank and matrix. The concentration of I.S. in the oil matrix is about 1 µg/kg. In the work presented in this application note, the I.S. working standard was added to g of the purified olive oil sample. The resulting I.S concentration in the matrix was µg/kg. Preparation of Working Standards (Olive Oil as Matrix) To prepare a 1 µg/ml stock standard solution, add 995 µl isopropanol, using a 1-mL pipette, and 5 µl of the 200 µg/ml standard solution, using a 10-µL pipette, to a 2-mL vial. The stock standard solution is used to prepare working standards as described in Table 5. Edible Oil Sample Preparation Prior to injection, filter oil through a 0.45-µm membrane (PTFE, Millex-LCR, Millipore). Table 1. Gradient Program for On-Line SPE Time Flow rate (ml/min) Solvent A (% vol.) Solvent B (% vol.) Solvent C (% vol.) Curve (%) Table 2. Gradient Program for Separation Time Flow rate Solvent A Solvent B Curve (%) (ml/min) (% vol.) (% vol.) Table 3. Valve Switching Programs for the Left and Right Valves Time (min) Left Valve Right Valve No Movement No Movement No Movement 61.5 No Movement 1-2 Time (min) Table 4. Wavelength Changes for RF2000 Fluorescence Detector Excitation Wavelength (nm) Emission Wavelength (nm) Determination of Polycyclic Aromatic Hydrocarbons (PAHs) in Edible Oils by Donor-Acceptor Complex Chromatography (DACC)-HPLC with Fluorescence Detection

60 Table 5. Preparation of the Working Standards (Oil as Matrix) Vial # Vial 1 Vial 2 (1.5 ml) Volume of 1 µg/ml PAH stock standard solution (µl) Volume of isopropanol (µl) Concentration of PAHs (µg/ml) Vial # (1.5 ml) Vial 3 Vial 4 Vial 5 Vial 6 Volume of Diluted Standard (Vial 1 or Vial 2) or Stock Standard (µl) Added weight of the cleaned olive oil used as matrix (containing I.S.) (g) Final concentration of PAHs (µg/kg) Final concentration of I.S. (µg/kg) 10 µl, Vial 1 10 µl, Vial 2 10 µl, stock standard 20 µl, stock standard Results and Discussion Description of the On-Line DACC-HPLC Method The flow scheme, shown in Figure 1, couples the DACC cleanup directly with the analytical HPLC run, using a second gradient pump and two column-switching valves. Figure 1 shows the valve positions at the time of the injection. The filtered and undiluted oil is injected directly, using isopropanol (IPA) to transfer the sample onto the enrichment column (DACC column). The analytical separation column is equilibrated with the second pump at the same time. After the analytes are bound to the DACC column, the right valve switches to flush out the oils and IPA, in a backflow mode, with acetonitrile/water (Figure 2). When all IPA and oils have been removed, the system switches the enrichment column into the analytical flow path (Figure 3). Pump A Autosampler Waste Enrichment Column Analytical Column Pump B Figure 1. Flow scheme for on-line sample preparation and analysis. The valves are positioned for injection of the sample on the enrichment column. Pump A Autosampler Figure 2. Isopropanol is flushed out of the enrichment column in backflow mode. Pump A Autosampler Waste Enrichment Column 4 5 Waste Enrichment Column Analytical Column Pump B Analytical Column Pump B Fluorescence Detector Fluorescence Detector Fluorescence Detector Figure 3. The enrichment column is switched into the analytical flow path, eluting the PAHs onto the analytical column for gradient separation followed by fluorescence detection. 60 Determination of Polycyclic Aromatic Hydrocarbons (PAHs) in Edible Oils by Donor-Acceptor Complex Chromatography (DACC)-HPLC with Fluorescence Detection

61 Reproducibility, Detection Limits, and Linearity Method reproducibility was estimated by making seven replicate injections of olive oil sample 1 spiked with the PAHs standard mix (vial #6 in Table 5) (Figure 4). Table 6 summarizes the retention time and peak area precision data. Calibration linearity for the determination of PAHs was investigated by making five replicate injections of a mixed standard of PAHs prepared at four different concentrations. The internal standard method was used to calculate the calibration curve and for real sample analysis. Table 7 reports the data from this determination as calculated by Chromeleon software. PAH method detection limits (MDLs) are also listed in Table 7. On-Line SPE Detection: Fluorescence Column: ChromSpher Pi, mm Sample: Mixture of PAHs standard Analytical Peaks: µg/kg Columns: Two Supelco PAH columns 1. Phenanthrene mm for separation 2. Anthracene 20 Eluent: A) water, 3. Fluoranthene 20 B) acetonitrile for both loading and 4. Pyrene 20 analysis pumps 5. Benzo[a]anthracene 20 C) isopropanol for loading pump 6. Chrysene 20 Flow Rate: 1 ml/min 7. Benzo[b]fluoranthene 20 Inj. Volume: 80 µl 8. Benzo[k]fluoranthene 20 Temperature: 30 ºC for column, 9. Benzo[a]pyrene ºC for autosampler 10. Dibenzo[a,h]anthracene Benzo[g,h,i]perylene Indeno[1,2,3-cd]pyrene Benzo[b]crysene (added as I.S.) 20 mv Figure 4. Overlay of chromatograms of seven serial injections of olive oil sample 1 spiked with a PAH standard mixture (20 µg/kg) Table 6. Reproducibility of Retention Times and Peak Areas a PAH RT RSD Area RSD Phenanthrene Anthracene Fluoranthene Pyrene Benzo[a]anthracene Chrysene Benzo[b]fluoranthene Benzo[k]fluoranthene Benzo[a]pyrene Dibenzo[a,h]anthracene Benzo[g,h,i]perylene Indeno[1,2,3-cd]pyrene a Seven injections of olive oil sample 1 spiked with 20 µg/kg mixed PAH standards. Table 7. Calibration Data for the 12 PAHs Phenols Equations r (%) MDL (µg/kg) Phenanthrene A = C Anthracene A = C Fluoranthene A = C Pyrene A = C Benzo[a]anthracene A = C Chrysene A = C Benzo[b]fluoranthene A = C Benzo[k]fluoranthene A = C Benzo[a]pyrene A = C Dibenzo[a,h]anthracene A = C Benzo[g,h,i]perylene A = C Indeno[1,2,3-cd]pyrene A = C The single-sided Student s test method (at the 99% confidence limit) was used for estimating MDL, where the standard deviation (SD) of the peak area of seven injections of olive oil sample 1 spiked with 2 µg/kg mixed PAHs standard is multiplied by 3.14 (at n = 7) to yield the MDL. 61 Determination of Polycyclic Aromatic Hydrocarbons (PAHs) in Edible Oils by Donor-Acceptor Complex Chromatography (DACC)-HPLC with Fluorescence Detection

62 Carryover Performance Carryover performance for the WPS 3000TSL autosampler was investigated by serial injections of 500 μg/kg of benzo[b]crysene (I.S.) and a purified olive oil sample prepared as a blank. Figure 5 shows exceptional carryover performance with external needle wash by acetonitrile both before and after the injection. There was no cross contamination observed when using the WPS 3000TSL autosampler for this application. On-Line SPE Column: ChromSpher Pi, mm Analytical Columns: Two Supelco PAH columns, mm for separation Eluent: A) water B) acetonitrile for both loading and analysis pumps C) isopropanol forloading pump 15 Flow Rate: 1 ml/min Inj. Volume: 80 µl Temperature: 30 ºC for column, 40 ºC for autosampler Detection: Fluorescence On-Line SPE Flow Rate: Column: ChromSpher Pi, mm Analytical Columns: Two Supelco PAH columns, mm for separation Detection: Eluent: A) water B) acetonitrile for both loading and analysis pumps C) isopropanol for loading pump 1,000 mv 100 A B 4.00 mv 1 ml/min Inj. Volume: 80 µl Temperature: 30 ºC for column, 40 ºC for autosampler Fluorescence Peaks: 1. Benzo[b]crysene (added as I.S.) A B Figure 5. Carryover test on the WPS-3000 autosampler. A) Purified olive oil spiked with 500 μg/kg of benzo[b]crysene (I.S.). B) Purified olive oil prepared as blank, analyzed immediately after A) mv 6 A B Figure 6. Overlay of chromatograms of A) untreated olive oil, and B) purified olive oil used as a blank Sample Analysis Two olive oil samples and one sesame oil sample were analyzed. The results are summarized in Table 8. Figure 7 shows chromatograms of the oil samples. Spike recoveries for these PAHs were in the ranges from 70% to 131%. Some PAHs were found in the edible oil samples. Five PAHs, phenanthrene, anthracene, benzo[a] anthracene, chrysene and benzo[a]pyrene, existed in all of the three samples and phenanthrene was obviously the most abundant PAH. 70 Effect of the Purified Olive Oil Used as Blank and Matrix One brand of olive oil was prepared as a blank and to serve as a matrix according to the procedure specified above. Figure 6 is an overlay of chromatograms of the original and purified olive oils, from which we observe that many ingredients were eliminated from the original olive oil. However, impurities persisted in the prepared olive oil used as a blank and matrix, which might have affected determination of some PAHs. To overcome this effect, the baseline of the purified olive oil blank was subtracted during data processing with Chromeleon software. 62 Determination of Polycyclic Aromatic Hydrocarbons (PAHs) in Edible Oils by Donor-Acceptor Complex Chromatography (DACC)-HPLC with Fluorescence Detection

63 On-Line SPE Column: ChromSpher Pi, mm Analytical Columns: Two Supelco PAH columns mm for separation Eluent: A) water, B) acetonitrile for both loading and analysis pumps C) isopropanol for loading pump Flow Rate: 1 ml/min Inj. Volume: 80 µl Temperature: 30 ºC for column, 40 ºC for autosampler 60 Detection: Fluorescence Peaks: 1. Phenanthrene 4. Pyrene 5. Benzo[a]anthracene 6. Chrysene 7. Benzo[b]fluoranthene 8. Benzo[k]fluoranthene 9. Benzo[a]pyrene 11. Benzo[g,h,i]perylene 12. Indeno[1,2,3-cd]pyrene On-Line SPE Column: ChromSpher Pi, mm Analytical Columns: Two Supelco PAH columns, mm for separation Eluent: A) water B) acetonitrile for both loading and analysis pumps C) isopropanol forloading pump Flow Rate: 1 ml/min Inj. Volume: 80 µl 120 Temperature: 30 ºC for column, 40 ºC for autosampler Detection: Fluorescence Sample: Mixture of PAHs standard 1 mv mv C B B A A Figure 7. Overlay of chromatograms of A) olive oil 1, B) olive oil 2, and C) sesame oil samples. Ruggedness of the SPE Column The tolerance of the SPE column used in this online analysis of PAHs in edible oils was investigated by comparing the separation of PAHs using two different SPE columns. One of these columns already had extracted over 600 injections of an edible oil sample; the other was nearly new. Figure 8 shows an overlay of chromatograms of PAHs using these two SPE columns. Final results of the PAH analyses are very similar, despite the different exposure levels of the two columns. Figure 8. Separation of PAHs in olive oil using different SPE columns. A) SPE column with 600 prior injections; B) new SPE column. Precautions Contaminants in solvents, reagents, glassware, and other sample processing hardware may cause method interferences, so glassware must be scrupulously cleaned. Use high-purity reagents and solvents to minimize interference problems. 63 Determination of Polycyclic Aromatic Hydrocarbons (PAHs) in Edible Oils by Donor-Acceptor Complex Chromatography (DACC)-HPLC with Fluorescence Detection

64 Table 8. Analytical Results for Olive Oil 1, Olive Oil 2, and Sesame Oil PAH Olive Oil 1 Olive Oil 2 Sesame Oil Detected (µg/kg) Added (µg/kg) Recovery (%) Detected (µg/kg) Detected (µg/kg) Phenanthrene Anthracene Fluoranthene ND ND Pyrene ND Benzo[a]anthracene Chrysene Benzo[b]fluoranthene ND 5 90 ND ND Benzo[k]fluoranthene ND 5 84 ND ND Benzo[a]pyrene Dibenzo[a,h]anthracene ND 5 84 ND ND Benzo[g,h,i]perylene ND 5 70 ND 1.2 Indeno[1,2,3-cd]pyrene ND 5 82 ND ND One sample and one spiked sample were prepared, and two injections of each were made. References 1) Commission Regulation (EC) No. 1881/2006 of 19 December 2006, Setting Maximum Levels for Certain Contaminants in Foodstuffs. Official Journal of the European Union 2006, 49, L364, ) Saravanabhavan, G.; Helferty, A.; Hodson, P. V.; Brown, R. S. A Multi-Dimensional High Performance Liquid Chromatographic Method for Fingerprinting Polycyclic Aromatic Hydrocarbons and Their Alkyl- Homologs in the Heavy Gas Oil Fraction of Alaskan North Slope Crude. J. Chromatogr., A 2007, 1156, ) Zuin, V. G.; Montero, L.; Bauer, C.; Popp, P. Stir Bar Sorptive Extraction and High-Performance Liquid Chromatography-Fluorescence Detection for the Determination of Polycyclic Aromatic Hydrocarbons in Mate Teas. J. Chromatogr., A 2005, 1091, ) Pino, V.; Ayala, J. H.; Afonso, A.M.; González, V. Determination of Polycyclic Aromatic Hydrocarbons in Seawater by High-Performance Liquid Chromatography with Fluorescence Detection Following Micelle-Mediated Preconcentration. J. Chromatogr., A 2002, 949, ) Bouvrette, P.; Hrapovic, S.; Male, K. B.; Luong, J. H. Analysis of the 16 Environmental Protection Agency Priority Polycyclic Aromatic Hydrocarbons by High Performance Liquid Chromatography-Oxidized Diamond Film Electrodes. J. Chromatogr., A 2006, 1103, ) Itoh, N.; Aoyagi, Y.; Yarita, T. Optimization of the Dopant for the Trace Determination of Polycyclic Aromatic Hydrocarbons by Liquid Chromatography/ Dopant-Assisted Atmospheric-Pressure Photoionization/Mass Spectrometry. J. Chromatogr., A 2006, 1131, ) Lintelmann, J.; Fischer, K.; Matuschek, G. Determination of Oxygenated Polycyclic Aromatic Hydrocarbons in Particulate Matter Using High- Performance Liquid Chromatography-Tandem Mass Spectrometry. J. Chromatogr., A 2006, 1133, ) Pandey, M.K.; Mishra, K.K.; Khanna, S.K.; Das, M. Detection of Polycyclic Aromatic Hydrocarbons in Commonly Consumed Edible Oils and Their Likely Intake in the Indian Population. J. Am. Oil Chem. Soc. 2004, 81, Determination of Polycyclic Aromatic Hydrocarbons (PAHs) in Edible Oils by Donor-Acceptor Complex Chromatography (DACC)-HPLC with Fluorescence Detection

65 9) Barranco, A.; Alonso-Salces, R. M.; Bakkali, A.; Berrueta, L. A.; Gallo, B.; Vicente, F.; Sarobe, M. Solid-Phase Clean-Up in the Liquid Chromatographic Determination of Polycyclic Aromatic Hydrocarbons in Edible Oils. J. Chromatogr., A 2003, 988, ) Bogusz, M. J.; El Hajj, S. A.; Ehaideb, Z.; Hassan, H.; Al-Tufail, M. Rapid Determination of Benzo(a)pyrene in Olive Oil Samples with Solid-Phase Extraction and Low-Pressure, Wide-Bore Gas Chromatography-Mass Spectrometry and Fast Liquid Chromatography with Fluorescence Detection. J. Chromatogr., A 2004, 1026, ) Barranco, A.; Alonso-Salces, R. M.; Corta, E.; Berrueta, L. A.; Gallo, B.; Vicente, F.; Sarobe, M. Comparison of Donor Acceptor and Alumina Columns for the Clean-Up of Polycyclic Aromatic Hydrocarbons from Edible Oils. Food Chem. 2004, 86, ) van Stijn, F.; Kerkhoff, M. A.; Vandeginste, B. G. Determination of Polycyclic Aromatic Hydrocarbons in Edible Oils and Fats by On-Line Donor-Acceptor Complex Chromatography and High-Performance Liquid Chromatography with Fluorescence Detection. J. Chromatogr., A 1996, 750, ) Maio, G.; Arnold, F.; Franz, H. Analysis of PAHs in Edible Oils by DACC-HPLC Coupling with Fluorescence Detection, Poster Presented at the 25th International Symposium on Chromatography, Paris, France, Oct. 4 8, LPN 1663, Dionex Corporation, Sunnyvale, CA. com/en-us/webdocs/7503_summit_x2_poster_pah_ application_isc04_sep042.pdf or search for PAH on 65 Determination of Polycyclic Aromatic Hydrocarbons (PAHs) in Edible Oils by Donor-Acceptor Complex Chromatography (DACC)-HPLC with Fluorescence Detection

66 Application Note 242 Robust and Fast Analysis of Tobacco-Specific Nitrosamines by LC-MS/MS INTRODUCTION Tobacco-specific nitrosamines (TSNA) are a group of carcinogens found only in tobacco products. They are formed from nicotine and related alkaloids during the production and processing of tobacco and tobacco products. 1 Due to their carcinogenic properties, efforts have been made to reduce TSNA levels in tobacco products. The desired goal of this investigation is to develop a sensitive, high-throughput method to monitor TSNA levels in tobacco and tobacco products. This study focuses on N'-nitrosonornicotine (NNN), N'-nitrosoanatabine (NAT), N-nitrosoanabasine (NAB), and 4-(methylnitrosamino)-1- (3-pyridyl)-1-butanone (NNK). Conventional methods for TSNA analysis are based on gas chromatography with a thermal energy analyzer (GC-TEA) 2,3, or high-performance liquid chromatography (HPLC) with various detection techniques such as UV and mass spectrometry (MS). 4 6 This application note (AN) describes a robust and fast LC-MS/MS method developed to analyze the four TSNAs in tobacco cigarettes after a simple sample preparation of ammonium acetate liquid extraction and filtration. The chromatographic separation of the four TSNAs was achieved within 3.5 min and these analytes were detected with great sensitivity and selectivity by tandem mass spectrometry using Selected Reaction Monitoring (SRM) detection. This method has been evaluated with respect to linearity, detection limits, precision, and accuracy. The ruggedness of the methodology was proven with more than 1000 replicate injections. EXPERIMENTAL Instrumentation HPLC P680 dual ternary pump ASI-100 autosampler TCC-100 thermostatted column compartment UVD340U detector Mass Spectrometer TSQ Quantum Access triple quadrupole mass spectrometer with heated electrospray ionization (HESI) interface. Software Xcalibur 2.0 Dionex DCMS Link for Xcalibur (version 2.0)* * DCMS Link is a Chromeleon -based software module providing the interface for controlling a wide range of Dionex chromatography instruments from different mass spectrometer software platforms. Conditions Chromatographic Conditions Analytical Column: Acclaim RSLC PA2 ( mm, 2.2 µm) Column Temperature: 60 C Injection Volume: 10 µl Mobile Phase: 10% CH 3 CN in buffer (1 mm NH 4 OAc, ph adjusted to ph 8.0 by NH 4 OH) Flow Rate: 0.50 ml/min Detection: UV at 230 nm TSQ Quantum Access SRM 66 Robust and Fast Analysis of Tobacco-Specific Nitrosamines by LC-MS/MS

67 Mass Spectrometric Conditions Needle Voltage: 2000 V Vaporizer Temp.: 350 C Sheath Gas: Auxiliary Gas: Ion Sweep Gas: 60 (arbitrary unit) 50 (arbitrary unit) 2 (arbitrary unit) Collision Gas Pressure: 1.5 mtorr Capillary Temp.: 350 C SRM Transitions: NNN m/z 11 V m/z 18 V NNK m/z 11 V m/z 38 V NNK-d m/z 11 V m/z 38 V NAT m/z 10 V m/z 5 V NAB m/z 10 V m/z 24 V The 1st SRM transition of each compound is used for quantification and the 2nd SRM transition is used for confirmation. Sample Preparation Tobacco cuts from five brands of cigarettes were weighed to 0.25 grams into 20 ml glass vials, and extracted with 10 ml of 100 mm ammonium acetate solution. The vials were placed on a swirl table and agitated for 30 min. The extracts were filtered through 0.25 µm membrane syringe filters. A 1.0 ml aliquot of each extract was then spiked with 10 µl internal standard (IStd, NNK-d 4 ) in preparation for LC-MS/MS analysis. RESULTS AND DISCUSSION Chromatography As shown in Figure 1, the four TSNAs are retained and completely resolved within 3.5 min, with the minimum of the retention factors (K' NNN ) greater than 4 and minimum resolution (R snat ) greater than 2.0. These demonstrated retention factors ensure analytes of interest are chromatographically separated from early eluted compounds that can suppress ionization. The total chromatographic resolution minimizes the possibility of ionization suppression and eliminates crosscontamination of SRM transitions between main analytes. Relative Abundance Absorbance (mau) NNN: 1.15 AA:437725, S/N:2394 NNK: 1.55 AA:493802, S/N:2223 NNK-d4: 1.51 AA: , S/N: NNN (Rs: 2.34) NNK (Rs: 4.29) NAT: 2.45 AA: , S/N:1465 NAB: 2.89 AA: , S/N:3105 NAB NAT (Rs: 2.05) NL: 7.13E NL: 6.66E NL: 1.73E NL: 1.42E NL: 4.77E MS/MS Conditions Needle Voltage: 2000 V Vaporizer Temp.: 350 C Sheath Gas: 60 (arbitrary unit) Auxiliary Gas: 50 (arbitrary unit) Ion Sweep Gas: 2 (arbitrary unit) Collision Gas Pre.: 1.5 mtorr Capillary Temp: 350 C SRM Transitions: Analyte SRM Collision Energy NNN V NNK V NNK-d V NAT V NAB V LC Conditions Column: Acclaim PA mm, 2.2 µm Mobile Phase: 10% CH 3 CN in buffer (1 mm NH 4 OAc, adjusted to ph = 8.0 with NH 4 OH) Column Temp.: 60 C Flow Rate: 0.5 ml/min (backpressure ~ 130 bar) Inj. Volume: 10 µl Analyte Conc.: 10 ppb each analyte 200 ppb for internal standard 5 ppm for UV detection Detection: UV at 230 nm or TSQ Quantum Access SRM with HESI Figure 1. Comparison of SRM and UV chromatograms of TSNAs on the Acclaim RSLC PA2 column. 67 Robust and Fast Analysis of Tobacco-Specific Nitrosamines by LC-MS/MS

68 Mass Spectrometry Tandem mass spectrometry provides specific and sensitive detection. The SRM chromatograms in Figure 1 show that quantification can be performed with a high level of confidence even at low concentration (10 ng/ml, 0.1 ng injected amount). The use of SRM detection enabled simplification of sample preparation and a significant reduction in total process time. Method Performance Internal standard quantification and calibration were performed using isotope labeled NNK-d 4 (IStd). Calibration standards were prepared at 8 levels from 2 ng/ml to 1000 ng/ml. Triplicate injections were performed to generate calibration curves, and a coefficient of determination (r 2 ) greater than 0.99 was achieved for each analyte. 1/X weighting provided best quantification for low concentration samples. Excellent linearity was achieved for NNK with r 2 = , suggesting that the use of isotope-labeled analogs as internal standards could provide better quantification accuracy and precision. The calibration curve is shown in Figure 2. Method detection limits (MDLs) were statistically calculated using the equation MDL = S t (99%, n=7) where S is the standard deviation, and t is student s t at 99% confidence interval. Seven replicate injections of a standard solution at 2 ng/ml were performed, and used for MDL calculations. The MDLs for TSNAs range from 0.22 ng/ml (NAB) to 0.38 ng/ml (NNK). Precision and accuracy were evaluated at 10 ng/ml and 100 ng/ml. Seven replicate injections of calibration standards at both levels were performed, and quantification was calculated based on the calibration curves. The results in Table 1 show that the quantification of NNK was very accurate and precise (< 4 % difference, < 4.5% RSD) at both low (10 ng/ml) and medium levels (100 ng/ml). However, significant inaccuracy was observed for NAB and NAT at the low level, showing 54.2% and 36.6% quantification differences respectively. This observation suggests better quantification accuracy may be achieved using an isotope labeled internal standard for each analyte for low-level TSNA analysis, for example, analyzing TSNAs in biological matrices. Area Ratio NNK Y = *X r 2 = W: 1/X NNK Y = *X r 2 = W: 1/X ppb Area Ratio ppb Table 1. Precision and Accuracy 10 ng/ml 100 ng/ml Accuracy (%) % RSD Accuracy (%) % RSD NAB NAT NNK NNN Figure 2. Calibration curve of NNK using NNK-d 4 as internal standard. 68 Robust and Fast Analysis of Tobacco-Specific Nitrosamines by LC-MS/MS

69 Method Ruggedness Since the analytical column was running under the very harsh conditions of high temperature and high ph mobile phase, method ruggedness was evaluated by repeated injections of standard solutions and tobacco extracts. As shown in Figure 3, chromatographic retention and resolution were well-maintained after more than 1000 injections. The column was cleaned with 50/50 CH 3 CN/H 2 O before injection 250 and a new inlet frit was installed before injection (The use of an in-line filter is highly recommended to prevent column clogging.) 70 NAB Determination of TSNAs in Tobacco Cigarettes Five brands of tobacco cigarettes were purchased from a local convenience store. Tobacco cuts from each brand of cigarettes were prepared (n = 3) following the procedure described previously. The results are shown in Table 2. The TSNA contents are significantly different between brands A and B and brands C, D, and E. The differences in TSNA content could be related to differences in the tobacco blended in each brand, such as type, origin, age of the tobacco, curing method, and storage conditions. The comparison of brands A and B (regular and light varieties of a premium US brand of cigarettes) shows no difference for TSNA contents except for a change in the amount of NAB. See Figure 4 for SRM chromatograms of TSNAs from brand A cigarette. Absorbance (mau) Absorbance (mau) Absorbance (mau) NNK (Rs: 4.29) NNN (Rs: 2.34) NNK (Rs: 4.09) NNN (Rs: 2.27) NNK (Rs: 3.68) NNN (Rs: 2.08) NAT (Rs: 2.05) NAT (Rs: 1.88) NAT (Rs: 1.63) NAB NAB Injection #1 Injection #250 Injection #1003 Relative Abundance NNN AA: , S/N: 1183 NNK AA: , S/N: 348 NNK-d AA: , S/N: 2018 NAT AA: , S/N: 798 NL: 2.28E NL: 4.57E NL: 6.29E NL: 2.08E Figure 3. Ruggedness of the Acclaim RSLC PA2 column NAB AA: , S/N: 95 NL: 6.45E Figure 4. SRM chromatograms of TSNAs from brand A cigarette. 69 Robust and Fast Analysis of Tobacco-Specific Nitrosamines by LC-MS/MS

70 Table 2. TSNAs in Tobacco Cigarettes* NAB NAT NNK NNN Note Brand A US brand, regular Brand B US brand, lights Brand C US brand, all natural Brand D International brand Brand E Asian brand *Amounts shown in Table 2 are ng/g tobacco cuts. CONCLUSION A rugged and ultrafast method for TSNA analysis was developed using HPLC-MS/MS on an Acclaim RSLC PA2 column. Four TSNAs were chromatographically retained and resolved within 3.5 min. Tandem mass spectrometry ensured selectivity and sensitivity. The use of isotope-labeled intenal standards for each analyte may improve the quantification accuracy, as was observed for NNK in this study. The ruggedness of the method was shown by more than 1000 injections of standards and tobacco extracts. The applicability of the method for the determination of TSNAs in tobacco cuts from five brands of cigarettes was demonstrated. REFERENCES 1. Hecht, S.S.; Hoffmann, D. Carcinogenesis 1988, 9, Adams, J.D.; Brunnemann, K.D.; Hoffmann, D. J. Chromatogr., A 1983, 256, Krull, I.S.; Swartz, M.; Hilliard, R.; Xie, K.H.; Driscoll, J.N. J. Chromatogr., A 1983, 260, Wu, W.; Ashley, D. L.; Watson, C. H. Anal. Chem. 2003, 75, Wagner, K.A.; Finkel, N.H.; Fossett, J.E.; Gillman, I.G. Anal. Chem. 2005, 77, Wu, J.; Joza, P.; Sharifi, M.; Rickert, W.S.; Lauterbach, J.H. Anal. Chem. 2008, 80, Robust and Fast Analysis of Tobacco-Specific Nitrosamines by LC-MS/MS

71 Application Note 270 Determination of Hydroxymethylfurfural in Honey and Biomass INTRODUCTION Hydroxymethylfurfural (HMF), or 5-hydroxymethyl- 2-furaldehyde, is a water-soluble heterocyclic organic compound derived from sugars. It is a derivative of furan and has both aldehyde and alcohol functional groups (Figure 1). Very low amounts of this compound are naturally found in fresh sugar-containing foods including milk, honey, fruit juices, spirits, and bread. Additionally, HMF is produced during food pasteurization and cooking as a result of dehydration of sugars such as glucose and fructose 1 and in the initial stages of the Maillard reaction, 2 a reaction between sugars and proteins responsible for changes in color and flavor of food. HMF is also formed during extended food storage under acidic conditions that favor its generation. Therefore, it is an indicator of excessive heat-treatment, spoilage, and of possible adulteration with other sugars or syrups. Although HMF is not yet considered a harmful substance, the National Institute of Environmental Health Sciences nominated HMF for toxicity testing 3 based on the potential for widespread exposure through consumed foods, and evidence for carcinogenic potential of other members of this class. As a result, many countries impose restrictions on maximum levels of HMF in food and beverages. 4 Beyond being an indicator of food quality, HMF is a biomass platform chemical because it can be used to synthesize a number of compounds that are currently derived from crude oil, including solvents, fuels, and monomers for polymer production (Figure 1). 5,6 HMF is readily derived from cellulose, either directly or through a two-step process involving hydrolysis and the formation of simple sugars. 7 Thus, it is important to measure HMF in matrices ranging from foods to treated cellulosic biomass. HO O 2-Hydroxymethylfuran O 2-Methylfuran O OH HO 2,5-Bis(Hydroxymethyl)furan HO HO O O O O O O OH O Levulinic acid OH 2,5-Furandicarboxylic acid 2-Hydroxymethylfuran O 2,5-Dimethylfuran O HO H Formic acid O 2,5-Dimethyltetrahydrolfuran Figure 1. HMF as precursor for a number of commercial chemicals. 71 Determination of Hydroxymethylfurfural in Honey and Biomass

72 There are spectrophotometric and HPLC methods available for HMF determination. One commonly used method is based on spectral absorbance at 284 nm. 8, 9 This direct-absorbance measurement could have interferences from other compounds present in the complex matrices. In the HPLC method, HMF is separated on a reversed-phase column, with water and methanol as the mobile phase, and then detected by UV absorbance. 10 This work describes a high-performance anionexchange chromatography with pulsed amperometric detection (HPAE-PAD)-based method for the determination of HMF in samples ranging from food (honey and pancake syrup) to treated biomass (corn stover and wood hydrolysate). A Dionex ICS-3000 system with a CarboPac PA1 column, electrolytically generated hydroxide eluent, and electrochemical detection with disposable Au-on-polytetrafluoroethylene (PTFE) working electrodes are used. The CarboPac PA1 is a highcapacity, rugged column suitable for determining monoand disaccharides, and has high resolution for HMF in a wide variety of matrices. The testing here demonstrates the linearity, limit of quantitation, limit of detection, precision, and recovery of HMF in diverse matrices ranging from honey to corn stover. It shows that PAD is an appropriate detection technique for HMF, with a broad linear range and low detection limit. Disposable electrodes provide short equilibration time and greater electrode-to-electrode reproducibility, compared to conventional electrodes. Compared to other disposable Au electrodes, the Au-on- PTFE electrodes have longer lifetime and can operate at higher hydroxide concentrations. The described method provides good sensitivity, consistent response, and can be routinely used for HMF analysis in foods and biomass applications, demonstrating the capability of HPAE-PAD for HMF determination in varied matrices. EQUIPMENT Dionex ICS-3000 or ICS-5000 system including: Gradient or Isocratic Pump, with the vacuum degas option installed EG Vacuum Degas Conversion Kit (Dionex P/N ) DC Detector/Chromatography Module 10 μl Injection loop Electrochemical Detector (P/N ) Carbohydrate PTFE Disposable Au Working Electrodes (P/N , package of 6) Ag/AgCl Reference Electrode (P/N ) 3 mil PTFE gaskets (P/N 63537) AS Autosampler Chromeleon Chromatography Data System (CDS) software Eluent Organizer, including 2 L plastic bottles and pressure regulator Polypropylene injection vials with caps (0.3 ml vial kit, P/N ) Nalgene 125 ml HDPE narrow mouth bottles (VWR P/N ) Nalgene 250 ml HDPE narrow mouth bottles (VWR P/N ) Nalgene 250 ml 0.2 µm nylon filter units (VWR P/N ) Nalgene 1000 ml 0.2 µm nylon filter units (VWR P/N ) REAGENTS AND STANDARDS Reagents Deionized (DI) water, Type I reagent grade, 18 MΩ-cm resistivity or better, filtered through a 0.2 μm filter immediately before use Standards HMF (Sigma Aldrich Cat # W501808) Fructose (Baker Analyzed Cat # M556-05) Xylose (Aldrich Chemical Company Cat # X-107-5) Sucrose (Sigma Cat # S-9378) Glucose (Sigma Cat # G-5250) Glycerol (JT Baker Cat # M778-07) Arabinose (Sigma Cat # A3131) 72 Determination of Hydroxymethylfurfural in Honey and Biomass

73 CONDITIONS Method Columns: Flow Rate: Inj. Volume: Temperature: 30 ºC Back Pressure: Eluent: Eluent Source: Detection: Background: CarboPac PA1 Analytical, mm (P/N ) CarboPac PA1 Guard, 4 50 mm (P/N 43096) 1.0 ml/min 10 μl (full loop) 2400 psi 50 mm KOH EGC II KOH with CR-ATC PAD nc Working Electrode: Carbohydrate PTFE Disposable Au Working Electrodes Reference Electrode: Mode: Ag/AgCl mode Noise: 30 pc Carbohydrate Waveform Time (s) Potential (V) Integration Begin End Reference electrode in Ag/AgCl mode PREPARATION OF SOLUTIONS AND REAGENTS Eluent Solutions Potassium Hydroxide (50 mm) Generate the potassium hydroxide (KOH) eluent online by pumping high-quality degassed, deionized (DI) water through the EGC II KOH cartridge. Chromeleon software tracks the amount of KOH used and calculates the remaining lifetime. Although electrolytic eluent generation delivers the best performance, manually prepared eluents can be used, if needed. For manually prepared eluent, use NaOH rather than KOH and prepare according to the general instructions for hydroxide eluents in Dionex Technical Note This method requires the installation of the ICS-3000 EG Vacuum Degas Conversion Kit (P/N ) to allow sufficient removal of the hydrogen gas formed with the potassium hydroxide eluent. 12 Stock Standard Solution Prepare a stock solution of 2 mg/ml HMF by dissolving 10 mg in 5 ml DI water in a plastic volumetric flask. Store aliquots of stock solutions in plastic containers at 4 C. Stock standards are stable for at least one month. Working Standard Solutions Prepare working standards at lower HMF concentrations by diluting appropriate volumes of the 2 mg/ml stock with deionized water. Prepare working standards daily. Store the standard solutions at <6 ºC when not in use. SAMPLE PREPARATION Honey and Syrup Prepare honey and syrup samples by dissolving 1 g in 100 ml of DI water and sonicating for 10 min. Store solutions in plastic containers at < 6 ºC. Further dilute syrup samples twofold with DI water before injection. Fructose Prepare a stock solution by dissolving 100 mg in 100 μl DI water. Dilute 500-fold with DI water before injection. Corn Stover and Wood Hydrolysate Centrifuge corn stover and wood hydrolysate samples at 14,000 rpm for 10 min and inject with DI water at a dilution of 1/1000 for analysis. Precautions Carryover can occur because the honey, syrup, and treated biomass samples have high concentrations of sugars such as glucose, xylose, and sucrose. A syringe flush of 1000 μl is recommended between samples. Column washes at 100 mm KOH are recommended if gradual retention time loss is observed. The application of 100 mm KOH changes the system equilibrium; re-equilibration at 50 mm for ~2 h is recommended to achieve high precision Determination of Hydroxymethylfurfural in Honey and Biomass

74 Column: CarboPac PA1, Analytical (4 250 mm) CarboPac PA1, Guard (4 50 mm) Eluent: 50 mm KOH Eluent Source: EGC II KOH Flow Rate: 1.0 ml/min Temperature: 30 C Inj. Volume: 10 µl (full loop) Detection: PAD (Au) Peaks: µg/ml 1. Glycerol 2. HMF Glucose 4. Xylose 5. Fructose 6. Sucrose A. Thermally stressed honey B. HMF standard Column: CarboPac PA1, Analytical (4 250 mm) CarboPac PA1, Guard (4 50 mm) Eluent: 50 mm KOH Eluent Source: EGC II KOH Flow Rate: 1.0 ml/min Temperature: 30 C Inj. Volume: 10 µl (full loop) Detection: PAD (Au) Peaks: µg/ml 1. Glycerol 2. HMF Glucose, Fructose 4. Sucrose nc nc A B Figure 2. HMF in thermally stressed honey. Figure 3. HMF in pancake syrup (high-fructose corn syrup). Table 1. Intraday and Between-Day Precisions for Honey and Corn Stover Samples Sample Amount (μg/ml) RT Precision (RSD) Peak Area Precision (RSD) Intraday Between-Day Intraday Between-Day Fresh honey 0.17 NC Thermally stressed honey Corn stover NC: Not Collected RESULTS AND DISCUSSION Figure 2 shows HMF in a thermally stressed honey sample. HMF elutes at 4.8 min and can be detected without interference from the other sugars. The HMF content in this sample was determined to be 330 mg/kg of honey. Typically, fresh honey has a low amount of HMF (<15 mg/kg). The HMF concentration increases as honey undergoes heat treatment to reduce viscosity and prevent crystallization to facilitate filling. 14 The EU Directive (110/2001) and the Codex Alimentarius (ALINORM 01/2000) standards limit HMF to 40 mg/kg for honey produced under European conditions and 80 mg/kg for honey coming from tropical countries. 4 In the fresh honey sample, HMF was determined to be 0.17 μg/ml (which amounts to 17 mg HMF/kg of honey, Table 1). The chromatogram of thermally stressed pancake syrup (Figure 3) shows the separation of HMF from other thermal degradation products. HMF is a product of thermal degradation of fructose, the main constituent of pancake syrup. HMF in high-fructose corn syrup (HFCS) is also a problem for beekeepers because they use HFCS as a source of sugar to feed bees when natural nectar sources are limited. Note that complex matrices like pancake syrup may have later-eluting peaks (e.g., pancake syrup has a peak at ~55 min, not shown), and the long retention time could interfere with subsequent injections if a shorter run time is used. 74 Determination of Hydroxymethylfurfural in Honey and Biomass

75 Column: CarboPac PA1, Analytical (4 250 mm) CarboPac PA1, Guard (4 50 mm) Eluent: 50 mm KOH Eluent Source: EGC II KOH Flow Rate: 1.0 ml/min Temperature: 30 C Inj. Volume: 10 µl (full loop) Detection: PAD (Au) Peaks: µg/ml 1. HMF Fructose Column: CarboPac PA1, Analytical (4 250 mm) CarboPac PA1, Guard (4 50 mm) Eluent: 50 mm KOH Eluent Source: EGC II KOH Flow Rate: 1.0 ml/min Temperature: 30 C Inj. Volume: 10 µl (full loop) Detection: PAD (Au) Traces: A. Treated corn stover B. HMF standard Peaks: µg/ml 1. Glycerol 2. Fucose 3. HMF Arabinose 5. Glucose 6. Xylose 7. Fructose nc 4 nc Figure 4. HMF in a thermally stressed fructose solution A 2 B Figure 5. HMF in acid-hydrolyzed corn stover. The described HPAE-PAD method was applied for HMF detection in fructose. The United States Pharmacopeia (USP) 15 and Food Chemicals Codex (FCC) have monographs 16 for the analysis of HMF in fructose and fructose injections. The USP monograph is based on the Seliwanoff test which depends on the reaction of HMF with resorcinol to form a red-colored compound, and the FCC monograph is a spectrophotometric method based on UV absorbance of HMF at 283 nm. HMF is present as an organic impurity in fructose and must be quantified and meet USP requirements before use as a food substance or as infusion fluids. HMF is formed during sterilization and storage. The chromatogram of thermally stressed fructose solution (Figure 4) shows the separation of HMF from other thermal degradation products. HMF is also formed in aqueous dextrose solutions and a HPAE-PAD-based method has been reported for quantification of HMF in commercial dextrose solutions. 17 Note that the thermally stressed honey, syrup, and fructose solutions were prepared by heating samples at 100±10 ºC for 4 h and cooling to room temperature before dilution and injection. Figures 5 and 6 show the chromatograms of acidhydrolyzed corn stover and a wood acid hydrolysate. HMF was detected without interference from the other sugars in both samples. 140 nc Columns: CarboPac PA1, Analytical (4 250 mm) CarboPac PA1, Guard (4 50 mm) Eluent: 50 mm KOH Eluent Source: EGC II KOH Flow Rate: 1.0 ml/min Temperature: 30 C Inj. Volume: 10 µl (full loop) Detection: PAD (Au) Figure 6. HMF in wood hydrolysate Peaks: 1. Glycerol 2. Fucose 3. HMF 4. Arabinose 5. Glucose 6. Xylose 7. Fructose The total run time for these samples was 15 min, providing high sample throughput, suggesting that this method can be used for online monitoring of HMF during biomass processing. 75 Determination of Hydroxymethylfurfural in Honey and Biomass

76 Analyte HMF Sample (Range μg/ml) Table 2. Linear Range and Precisions for HMF Standards Corr. Coeff. (r 2 ) RT (min) Concentration Used for Precision Injections (μg/ml) RT Precision (RSD) Peak Area (nc*min) Peak Area Precision (RSD) Honey (0.1 50) Corn Stover ( ) Table 3. Recoveries of HMF in Spiked Honey and Corn Stover Samples Analyte Sample Amount Found (μg/ml) HMF Amount Added (μg/ml) Average Recovery ([%]n = 3 days) Thermally stressed honey Corn stover Linear Range, Limit of Quantitation, Limit of Detection To determine the linearity of the method, calibration standards were injected in triplicate covering the expected concentration range of HMF in food and biomass samples. Calibration plots produced correlation coefficient (r 2 ) values of in the range 0.1 to 50 μg/ml for food applications and in the range 0.5 to 1000 μg/ml for biomass applications (Table 2). A least squares regression fit with weighting was used to accurately represent the lower values of the calibration curve. The limit of detection (LOD) was determined by measuring the peak-to-peak noise in a representative one-minute segment of baseline where no peaks elute, followed by analyzing a standard at a concentration expected to provide a chromatogram with a signalto-noise (S/N) ratio of 3. Similarly, the lower limit of quantitation (LOQ) was determined by injecting a standard at a concentration that resulted in a S/N ratio of 10. Typical baseline noise for this method was 20 to 40 pc. The LOD and LOQ for this method were 0.04 μg/ml and 0.10 μg/ml, respectively. Precision The peak area and retention time (RT) precisions were determined for seven replicate injections of an HMF standard. The concentrations used for precision injections were 0.5 μg/ml for food applications and 5.0 μg/ml for biomass applications (Table 2). The retention time precisions (RSD) for the two concentrations were 0.12 and 0.10, respectively. The corresponding peak area precisions were 0.5 and The high retention time precisions were attributed to consistent generation of high-purity KOH using the eluent generator. Intraday and between-day precisions for HMF in honey and corn stover were evaluated over three consecutive days. The RT precisions were in the range of 0.06 to 0.67%, and peak area precisions were in the range of 0.33 to 3.13%. The high precisions suggest that this method can be used to measure HMF in complex matrices. Accuracy The accuracy of the method was verified by determining recoveries of HMF in spiked honey and acid-hydrolyzed corn stover samples over three consecutive days. The amount of HMF in a fresh honey sample was 0.17 μg/ml (Table 1: this equates to 17 mg of HMF per kg of honey). The thermally stressed honey had 3.4 μg/ml HMF (Figure 2), and was spiked with 2.9 μg/ml HMF. The treated corn stover sample (at 1000-fold dilution) had 4 μg/ml HMF and was spiked with 5.3 μg/ml HMF. Recoveries were calculated from the difference in response between the spiked and unspiked samples. Intraday concentration RSD was 1.7% for both honey and corn stover. The average recovery of HMF in honey was 103% and in corn stover was 112% (Table 3). 76 Determination of Hydroxymethylfurfural in Honey and Biomass

77 CONCLUSION This study describes a HPAE-PAD method for the accurate determination of HMF in foods like honey and in biomass like acid-hydrolyzed corn stover. The method uses the CarboPac PA1 column with electrolytically generated hydroxide eluent. The method is shown to have a broad linear range, high precisions, and low detection limits. The disposable Au working electrode provides consistently high detector response, assuring greater instrument-to-instrument and lab-to-lab reproducibility. This configuration needs only addition of deionized water for continuous operation. In summary, the described HPAE-PAD-based HMF analysis method is accurate and reliable, and should be applicable to online monitoring of HMF levels in food and biomass applications. SUPPLIERS VWR, 1310 Goshen Parkway, West Chester, PA 19380, U.S.A., Tel: Sigma-Aldrich Chemical Co., P.O. Box 2060, Milwaukee, WI 53201, U.S.A., Tel: REFERENCES 1. Belitz, H. D.; Grosch, W. Food Chemistry; Springerverlag: Berlin, 2009; pp and Chichester, C. O. Advances in Food Research. In Advances in Food and Nutrition Research; Academic Press: Boston, 1986; pp Federal Register: Intent to Conduct Toxicological Studies; cfm?objectid=06f48e77-a3e8-414d- D2C0AD55F7AEA8FA (accessed Dec 15, 2010). 4. Codex Alimentarius, Alinorm 01/25 (2000), Draft Revised Standard for Honey at Step 8 of the Codex Procedure; EU Directive /1/110/2001 of 02/12/2001 (L 10/47). 5. Kunz, M. Hydroxymethyl Furfural, a Possible Basic Chemical Industrial Intermediate, Studies in Plant Science, Inulin and Inulin Containing Crops 1993, 3, Larousse, C., Rigal, L.; Gaset, A. Synthesis of 5,5/-oxydimethylene bis (2-furfural) by Thermal Dehydration of 5-Hydroxymethyl Furfural in the Presence of Dimethylsulfoxide. J. Chem. Technol. Biotechnol 1992, 53, Sua, Y.; Browna, H. M.; Huanga, X.; Zhoua, X.; Amonettea, J. E.; Zhang, Z. C. Single-Step Conversion of Cellulose to 5-hydroxymethylfurfural (HMF), a Versatile Platform Chemical. Appl. Catal., A 2009, 361 (1-2) (June 20), White, J. Spectrophotometric Method for Hydroxymethyl Furfural in Honey. Journal of the Association of Official Analytical Chemists 1979, 62, Winkler, O., Beitrag zum Nachwals und zur Bestimmung von Oxymethylfurfural in Honig and Kunsthonig. Zeitschrift für Lebensmittel Untersuchung und Forshung 1955, 102(3), Zappalà, M.; Fallico, B.; Arena. E.; Verzera, A. Methods for the Determination of HMF in Honey: a Comparison. Food Control 16(3), Dionex Corporation, Eluent Preparation for High- Performance Anion-Exchange Chromatography with Pulsed Amperometric Detection. Technical Note 71, LPN , 2009, Sunnyvale, CA. 12. ICS-3000 EG Vacuum Degas Conversion Kit Installation Instructions; Document No ; Dionex Corporation; 2005, Sunnyvale, CA. 13. Dionex Corporation, Determination of Tobramycin and Impurities Using HPAE-PAD. Application Note 61, LPN 1626, 2004, Sunnyvale, CA. 14. Subramanian, R.; Umesh Hebbar, H.; Rastogi, N. K. Processing of Honey: A Review. International Journal Food Prot. 2007, 10(1), United States Pharmacopeia (USP), USP 34-NF 29, Food Chemicals Codex (FCC), FCC 7, Xu, H.; Templeton, A. C.; Reed, R. A. Quantification of 5-HMF and Dextrose in Commercial Aqueous Dextrose Solutions. J. Pharm. Biomed. Anal. 2003, 32, Determination of Hydroxymethylfurfural in Honey and Biomass

78 Column Selection Guide

79 Silica Columns Reversed-Phase (RP) Mixed-Mode HILIC Application-Specific Acclaim 120 C18 Acclaim 120 C8 Acclaim 300 C18 Acclaim Polar Advantage (PA) Acclaim Polar Advantage II (PA2) Acclaim Phenyl-1 Acclaim Trinity P1 Acclaim Mixed-Mode WAX-1 Acclaim Mixed-Mode WCX-1 Acclaim Mixed-Mode HILIC-1 Acclaim HILIC-10 Acclaim Organic Acid Acclaim Surfactant Acclaim Explosives E1 Acclaim Explosives E2 Acclaim Carbamate Example Applications General Applications Specific Applications Neutral Molecules Anionic Molecules Cationic Molecules Amphoteric/ Zwitterionic Molecules Mixtures of Neutral, Anionic, Cationic Molecules Surfactants Organic Acids Environmental Contaminants Vitamins Pharmacutical Counterions High hydrophobicity Fat-soluble vitamins, PAHs, glycerides Intermediate hydrophobicity Steroids, phthalates, phenolics Low hydrophobicity Acetaminophen, urea, polyethylene glycols High hydrophobicity NSAIDs, phospholipids Intermediate hydrophobicity Asprin, alkyl acids, aromatic acids Low hydrophobicity Small organic acids, e.g. acetic acids High hydrophobicity Antidepressants Intermediate hydrophobicity Beta blockers, benzidines, alkaloids Low hydrophobicity Antacids, pseudoephedrine, amino sugars High hydrophobicity Phospholipids Intermediate hydrophobicity Amphoteric surfactants, peptides Low hydrophobicity Amino acids, aspartame, small peptides Neutrals and acids Artificial sweeteners Neutrals and bases Cough syrup Acids and bases Drug active ingredient with counterion Neutrals, acids, and bases Combination pain relievers Anionic SDS, LAS, laureth sulfates Cationic Quats, benzylalkonium in medicines Nonionic Triton X-100 in washing tank Amphoteric Cocoamidopropyl betaine Hydrotropes Xylenesulfonates in handsoap Surfactant blends Noionic and anionic surfactants Hydrophobic Aromatic acids, fatty acids Hydrophilic Organic acids in soft drinks, pharmaceuticals Explosives U.S. EPA Method 8330, 8330B Carbonyl compounds U.S. EPA 1667, 555, OT-11; CA CARB 1004 Phenols Compounds regulated by U.S. EPA 604 Chlorinated/Phenoxy acids U.S. EPA Method 555 Triazines Compounds regulated by U.S. EPA 619 Nitrosamines Compounds regulated by U.S. EPA 8270 Benzidines U.S. EPA Method 605 Perfluorinated acids Dionex TN73 Microcystins ISO Isocyanates U.S. OSHA Methods 42, 47 Carbamate insecticides U.S. EPA Method Water-soluble vitamins Vitamins in dietary supplements Fat-soluble vitamins Vitamin pills Anions Inorgaic anions and organic acids in drugs Cations Inorgaic cations and organic bases in drugs Mixture of Anions and Cations Screening of pharmaceutical counterions API and counterions Naproxen Na + salt, metformin Cl - salt, etc. Column Selection Guide and Specifications 79

80 Polymer Columns IonPac AS23 IonPac AS22 IonPac AS22-Fast IonPac AS14/A IonPac AS12A IonPac AS9/HC/SC IonPac AS4A/SC IonSwift MAX-100 IonPac AS24 IonPac AS21 IonPac AS20 IonPac AS19 IonPac AS18 IonPac AS18-Fast IonPac AS17-C IonPac AS16 IonPac AS15 IonPac AS11(-HC) IonPac AS10 IonPac AS7 IonPac AS5 IonPac Fast Anion IIIA OmniPac PAX-100 OmniPac PAX-500 IonPac CS18 IonPac CS17 IonPac CS16 IonPac CS15 IonPac CS14 IonPac CS12A IonPac CS11 IonPac CS10 IonPac CS5A OmniPac PCX-100 OmniPac PCX-500 AminoPac PA10 AminoPac PA1 CarboPac PA200 CarboPac PA100 CarboPac PA20 CarboPac PA10 CarboPac PA1 CarboPac MA1 DNAPac PA200 DNAPac PA100 ProPac WAX/SAX ProPac WCX/SCX ProPac IMAC ProPac HIC ProPac PA1 ProSwift IonPac ICE-AS6 IonPac ICE-AS1 IonPac ICE-Borate IonPac NS1 MODE ORGANIC MOLECULES BIO-MOLECULES CATIONS ANIONS Inorganic Anions Oxyhalides Bromate Perchlorate Organic Acids Phosphoric/Citric Acids Poly/High-Valence Anions Hydrophobic Anions Hydrophobic/Halogenated Anions Anionic Neutral Molecules Inorganic Cations Sodium/Ammonium Amines/Polyvalent Amines Aliphatic/Aromatic Amines Alkanol/Ethhanolamines Biogenic Amines Transition/Lanthanide Metals Hydrophobic Cations Cationic Neutral Molecules Amino Acids Phosphorylated Amino Acids Amino Sugars Oligosccharides Mono-/Di-Saccharides Glycoproteins Alditols/Aldoses mono/di Saccharides ds Nucleic Acids Single-Stranded Oligonucleotides Peptides Proteins Metal-binding Proteins Monoclonal antibodies Aliphatic Organic Acids Alcohols Borate Large Molecules, Anions Small Molecules Small Molecules/LC-MS Polar/Non-Polar Small Molecules Hydrophobic/Aliphatic Organic Acids Surfactant Formulations Explosives/EPA 8330 Anion Exchange / Carbonate Anion Exchange / Hydroxide Cation Exchange Multi-Mode Affinity Ion Exclusion Reversed Phase Anion Exchange/Other 80 Column Selection Guide and Specifications Column Selection Guide and Specifications 81

81 Column Specifications IC Anion Columns Column IonPac AS24 IonPac AS23 IonPac AS22 IonPac AS21 IonPac AS20 IonPac AS19 IonPac AS18 IonPac AS17-C IonPac AS16 IonPac AS15 IonPac AS15-5mm Format Primary Eluent Application mm Hydroxide Recommended column for haloacetic acids prior to MS or MS/MS detection mm mm mm mm Carbonate Carbonate Recommended column for inorganic anions and oxyhalides. Trace bromate in drinking water. Recommended column for fast analysis of common inorganic anions mm Hydroxide Recommended column for trace perchlorate prior to MS or MS/MS detection mm mm mm mm mm mm mm mm mm mm mm mm Hydroxide Hydroxide Hydroxide Hydroxide Hydroxide Hydroxide Recommended column for trace perchlorate prior to suppressed conductivity detection. Recommended column for inorganic anions and oxyhalides. Trace bromate in drinking water. Recommended column for the analysis of common inorganic anions. Trace anions in HPW matrices. Carboxylated resin, no sulfate blank. Low capacity for fast analysis of common inorganic anions using gradient elution with the Eluent Generator. High capacity for hydrophobic anions including iodide, thiocyanate, thiosulfate, and perchlorate. Polyvalent anions including: polyphosphates and polycarboxylates High capacity for trace analysis of inorganic anions and low molecular weight organic acids in high purity water matrices mm Hydroxide Fast run, high capacity for trace analysis of inorganic anions and low molecular weight organic acids in high purity water matrices. Particle Diameter Substrate Crosslinking Latex Diameter Latex Crosslinking Capacity (per column) Functional Group 7 µm 55% µeq Alkanol quaternary ammonium 6 µm 55% µeq 320 µeq 6.5 µm 55% µeq 210 µeq Alkyl quaternary ammonium Alkyl quaternary ammonium 7.0 µm 55% µeq Alkanol quaternary ammonium 7.5 µm 55% µeq 310 µeq 7.5 µm 55% µeq 350 µeq 7.5 µm 55% 65 nm 10.5 µm 55% 75 nm 9 µm 55% 80 nm 8% 75 µeq 285 µeq 6% 7.5 µeq 30 µeq 1% 42.5 µeq 170 µeq 9 µm 55% µeq 225 µeq Alkanol quaternary ammonium Alkanol quaternary ammonium Alkanol quaternary ammonium Alkanol quaternary ammonium Alkanol quaternary ammonium Alkanol quaternary ammonium 5 µm 55% µeq Alkanol quaternary ammonium Hydrophobicity Ultralow Ultralow Ultralow Ultralow Ultralow Low Low Low Ultralow Medium- High Medium- High Column IonPac AS12A IonPac AS11-HC IonPac AS11 IonPac AS10 IonPac AS9-HC IonPac AS9-SC IonPac AS4A-SC IonPac Fast Anion IIIA IonPac AS7 IonPac AS5A IonPac AS5 Format mm mm mm mm mm mm mm mm mm mm Primary Eluent Carbonate Hydroxide Hydroxide Hydroxide Carbonate Application Moderate capacity for analysis of inorganic anions and oxyhalides. Trace chloride and sulfate in high carbonate matrices. High capacity for the determination of organic acids and inorganic anions in uncharacterized samples. Low capacity for fast profiling of organic acids and inorganic anions in well-characterized samples. High capacity for the analysis of inorganic anions and organic acids in high nitrate samples. High-capacity column for inorganic anions and oxyhalides. Trace bromate in drinking water mm Carbonate Low capacity for fast analysis of inorganic anions and oxyhalides. Specified column in US EPA Method (B) mm mm Carbonate Low capacity for fast analysis of common inorganic anions. Specified column in U.S. EPA Method (A) mm Hydroxide Recommended column for phosphoric and citric acids in cola soft drinks mm Specialty Eluents Polyvalent anions including chelating agents, polyphosphates and polyphosphonates. Cyanide, sulfide, hexavalent chromium, and arsenic speciation mm Hydroxide Low capacity for fast profiling of organic acids and inorganic anions in well-characterized samples mm Hydroxide Metal-EDTA complexes, metalcyanide complexes, and oxyanions. Particle Diameter Substrate Crosslinking Latex Diameter 9 µm 55% 140 nm 9 µm 55% 70 nm 13 µm 55% 85 nm 8.5 µm 55% 65 nm 9 µm 55% 90 nm 13 µm 55% 110 nm 13 µm 55% 160 nm Latex Crosslinking Capacity (per column) 0.20% 13 µeq 52 µeq 6% 72.5 µeq 290 µeq 6% 11 µeq 45 µeq 5% 42.5 µeq 170 µeq 18% 48 µeq 190 µeq Functional Group Alkyl quaternary ammonium Alkanol quaternary ammonium Alkanol quaternary ammonium Alkyl quaternary ammonium Alkyl quaternary ammonium 20% µeq Alkyl quaternary ammonium 0.50% 5 µeq 20 µeq Alkanol quaternary ammonium 7.5 µm 55% µeq Alkanol quaternary ammonium 10 µm 2% 530 nm 5 µm 2% 60 nm 15 µm 2% 120 nm 5% 100 µeq Alkyl quaternary ammonium 4% 35 µeq Alkanol quaternary ammonium 1% 20 µeq Alkanol quaternary ammonium Hydrophobicity Medium Medium- Low Very Low Low Medium- Low Medium- Low Medium- Low Ultralow Medium- High Low Low IonPac AS14A- 5 µm mm Carbonate Recommended column for fast analysis of common inorganic anions. 5 µm 55% ueq Alkyl quaternary ammonium Medium IonPac AS14A mm Carbonate For analysis of common inorganic anions. 7 µm 55% µeq Alkyl quaternary ammonium Medium IonPac AS mm mm Carbonate Moderate capacity for fast analysis of common inorganic anions. 9 µm 55% µeq 65 µeq Alkyl quaternary ammonium Medium- High 82 Column Selection Guide and Specifications Column Selection Guide and Specifications 83

82 IC Cation Columns Ion-Exclusion Columns Column IonPac CS18 IonPac CS17 IonPac CS16 IonPac CS15 IonPac CS14 IonPac CS12A- MS IonPac CS12A- 5 µm IonPac CS12A IonPac CS11 IonPac CS10 IonPac CS5A Format Primary Eluent Application mm MSA Recommended column for polar amines (alkanolamines and methylamines) and moderately hydrophobic and polyvalent amines (biogenic and diamines). Nonsuppressed mode when extended calibration linearity for ammonium and weak bases is required mm mm mm mm mm mm mm mm MSA MSA MSA MSA Recommended column for hydrophobic and polyvalent amines (biogenic amines and diamines) Recommended column for disparate concentration ratios of adjacenteluting cations such as sodium and ammonium. Can be used for alkylamines and alkanolamines. Disparate concentration ratios of ammonium and sodium. Trace ethanolamine in high-ammonium or high- potassium concentrations. Alkanolamines. Aliphatic amines, aromatic amines, and polyamines plus mono- and divalent cations mm MSA IC-MS screening column for fast elution and low flow rates required for interfacing with IC-MS mm MSA Recommended column for high efficiency and fast analysis (3 min) of mono- and divalent cations mm mm MSA Recommended column for the separation of mono- and divalent cations. Manganese morpholine, alkylamines, and aromatic amines mm HCl + DAP Separation of mono- and divalent cations. Ethanolamines if divalent cations are not present mm HCl + DAP Separation of mono- and divalent cations mm mm Pyridine dicarboxylic acid Recommended column for transition and lanthanide metals analysis. Aluminum analysis. Particle Diameter Substrate Crosslinking Latex Diameter Latex Crosslinking Capacity (per column) Functional Group 6 µm 55% µeq Carboxylic acid 7 µm 55% µeq 1.45 µeq 5 µm 55% µeq 8.4 µeq 8.5 µm 55% µeq 2.8 µeq 8.5 µm 55% µeq 1.3 µeq Carboxylic acid Carboxylic acid Carboxylic acid/ phosphonic acid/ crown ether Carboxylic acid 8.5 µm 55% µeq Carboxylic acid/ phosphonic acid 5 µm 55% µeq Carboxylic acid/ phosphonic acid 8.5 µm 55% µeq 2.8 µeq Carboxylic acid/ phosphonic acid Hydrophobicity Medium Very Low Medium Medium Low Medium Medium Medium 8 µm 55% 200 nm 5% µeq Sulfonic acid Medium 8.5 µm 55% 200 nm 5% 0.08 µeq Sulfonic acid Medium 9 µm 55% 140 nm 75 nm 10% 20% 0.02 µeq/ µeq 0.04 µeq/ 0.01 µeq Sulfonic acid/ alkanol quaternary ammonium - Column Format Primary Eluent Application IonPac ICE-AS1 IonPac ICE-AS6 IonPac ICE- Borate mm mm Heptafluorobutyric acid mm Heptafluorobutyric acid Organic acids in high ionic strength matrices. Fast separation of organic acids. Organic acids in complex or high ionic strength matrices mm MSA/ Mannitol Trace concentrations of borate Acclaim General and Specialty Columns Column Bonded Phase USP Type Endcapped Substrate Particle Diameter Substrate Crosslinking Latex Diameter Latex Crosslinking Capacity (per column) 7.5 µm 8% µeq 27 µeq Functional Group Sulfonic acid Hydrophobicity Ultra Low 8 µm 8% µeq Sulfonic and carboxylic acid Moderate 7.5 µm 8% µeq Sulfonic acid Ultra Low Particle Shape Particle Size Metal Impurity (ppm) Na, Fe, AL Average Pore Diameter Surface Area (m 2 /g) Mixed-Mode WAX Proprietary na Proprietary 5 µm 120 Å 300 na alkyl amine Mixed-Mode HILIC Proprietary na Proprietary 5 µm 120 Å 300 na alkyl diol Mixed-Mode WCX Proprietary na Proprietary 5 µm 120 Å 300 na alkyl carboxyl Organic Acid (OA) Proprietary na Yes 5 µm 120 Å % Surfactant and Proprietary na Yes 5 µm 120 Å 300 na Explosives E1/2 120 C18 C18 L1 Yes 2, 3 and 5 µm 120 Å % 120 C8 C8 L7 Yes Ultrapure Spherical 3 and 5 µm <10 ppm 120 Å % silica 300 C18 C18 L1 Yes 3 µm 300 Å 100 7% Polar Advantage Sulfamido C16 na Yes 3 and 5 µm 120 Å % Polar Advantage II Amide C18 na Yes 2, 3 and Å % µm HILIC Proprietary Yes 3 µm 120 Å 300 hydrophilic Phenyl-1 Proprietary alkyl phenyl Yes 3 µm 120 Å 300 Carbamate Proprietary Yes 3 and 5 µm 120 Å 300 alkyl group Trinity Yes 120 Å 300 Total Carbon Content 84 Column Selection Guide and Specifications Column Selection Guide and Specifications 85

83 Bio Columns Protein Column Phase Target Applications MAbPac SEC-1 MAbPac SCX-10 Base Matrix Material Substrate Crosslinking Capacity Recommended Flow Rate Solvent Compatibility Maximum Backpressure ph Range Column Phase Target Applications ProSwift WAX-1S Weak Anion Exchange Fast protein separation with good resolution using Anion Exchange Base Matrix Material Monolith; polymethacrylate with tertiary amine (DEAE) functional group Substrate Crosslinking Monolith Standard permeability Capacity 18 mg/ml BSA Recommended Flow Rate ml/min (4.6 mm), (1.0 mm) Solvent Compatibility Most common organic solvents Maximum Backpressure 1000 psi (4.6 mm) 2000 psi (1.0 mm) ph Range ProPac WCX-10 ProPac SCX-10 ProPac SCX-20 ProPac WAX-10 ProPac SAX-10 Weak Cation Exchange Strong Cation Exchange Weak Anion Exchange Strong Anion Exchange High resolution and high efficiency separations of proteins and glycoproteins, pi =3-10, MW>10,000 units High resolution and high efficiency separations of proteins and glycoproteins, pi =3-10, MW>10,000 units High resolution and high efficiency separations of proteins and glycoproteins, pi =3-10, MW>10,000 units High resolution and high efficiency separations of proteins and glycoproteins, pi =3-10, MW>10,000 units 10-µm diameter nonporous substrate to which is grafted a polymer chain bearing carboxylate groups. 10 µm diameter nonporous substrate to which is grafted a polymer chain bearing sulfonate groups. 10 µm diameter non-porous substrate to which is grafted a polymer chain bearing tertiary amine groups. 10 µm diameter nonporous substrate with grafted polymer chain bearing quaternary ammonium groups. 55% 6 mg/ ml lysozyme 55% 3 mg/ ml lysozyme 55% 5 mg/ ml BSA/ ml 55% 15 mg/ ml BSA ml/min ml/min ml/min ml/min 80% ACN, acetone. Incompatable with alcohols and MeOH 80% ACN, acetone, MeOH 80% ACN, acetone, MeOH, 80% ACN, acetone, MeOH 3000 psi (21 MPa) 3000 psi (21 MPa) 3000 psi (21 MPa) 3000 psi (21 MPa) ProSwift WCX-1S ProPac IMAC-10 ProSwift ConA-1S ProPac HIC-10 Weak Cation Exchange Immobilized Metal Affinity Reversed- Phase Fast protein separation with good resolution using Cation Exchange High resolution separation of certain metal-binding proteins and peptides Protein separation using hydrophobic interaction with salt gradient elution Monolith; polymethacrylate with carboxylic acid (CM) functional group 10 µm diameter nonporous polystyrene divinylbenzene substrate with poly (IDA) grafts. Spherical 5 µm, ultrapure silica, 300 A, surface area 100 m 2 / g, Monolith Standard permeability 23 mg/ml Lysozyme 55% >60 mg lysozyme/ ml gel (4 x 250 mm) n/a 340 mg lysozyme per 7.8 x 75 mm column ml/min (4.6 mm), (1.0 mm) 1.0 ml/min 1.0 ml/ min Most common organic solvents EtOH, urea, NaCl, non- ionic detergents, glycerol, acetic acid, guanidine HCl 2M Ammonium sulfate/ phosphate salts, organic solvent for cleanup 1000 psi (4.6 mm) 2000 psi (1.0 mm) 3000 psi (21MPa) ,000 psi ProSwift RP-1S Reversed- Phase Fast protein separation with high capacity using Reversed Phase Monolith; polystyrenedivinylbenzene with phenyl functional group Monolith Standard permeability 5.5 mg/ml Insulin 2 4 ml/min Most common organic solvents 2800 psi (19.2 Mpa) 1 14 ProSwift RP-2H Reversed- Phase Fast protein separation with high capacity using Reversed Phase Monolith; polystyrenedivinylbenzene with phenyl functional group Monolith High permeability 1.0 mg/ml Lysozyme 1 10 ml/min Most common organic solvents 2800 psi (19.3 Mpa) 1 14 ProSwift RP-4H ProSwift RP-3U Reversed- Phase Fast protein separation with high capacity using Reversed Phase Monolith; polystyrenedivinylbenzene with phenyl functional group Monolith Ultrahigh permeability 0.5 mg/ml Lysozyme 1 16 ml/min Most common organic solvents 2800 psi (19.3 Mpa) 1 14 ProSwift SAX-1S Strong Anion Exchange Fast protein separation with good resolution using Anion Exchange Monolith; polymethacrylate with quaternary amine functional group Monolith Standard permeability 18 mg/ml BSA (4.6 mm), (1.0 mm) Most common organic solvents 1000 psi (4.6 mm) 2000 psi (1.0 mm) ProSwift SCX-1S Strong Cation Exchange Fast protein separation with good resolution using Cation Exchange Monolith; polymethacrylate with sulfonic acid fuctional group Monolith Standard permeability 30 mg/ml Lysozyme ml/min (4.6 mm) Most common organic solvents 1000 psi (4.6 mm) Column Selection Guide and Specifications Column Selection Guide and Specifications 87

84 Carbohydrate Column Target Applications Base Matrix Material Substrate Crosslinking Latex Crosslinking Capacity Recommended Eluents Recommended Flow Rate Solvent Compatibility Maximum Backpressure ph Range CarboPac MA1 Reduced monoand disaccharide analysis. 7.5 µm diameter macroporous substrate fully functionalized with an alkyl quaternary ammonium group 15% No latex 1450 µeq (4 250 mm) Hydroxide 0.4 ml/min 0% 2000 psi (14 MPa) 0 14 CarboPac PA1 General purpose mono-, di-, and oligosaccharide analysis 10 µm diameter nonporous substrate agglomerted with a 500 nm MicroBead quaternary ammonium functionalized latex 2% 5% 100 µeq (4 250 mm) Hydroxide, acetate/ hydroxide 1.0 ml/min 0 5% 4000 psi (28 MPa) 0 14 CarboPac PA10 Monosaccharide compositonal anaylysis 10 µm diameter nonporous substrate agglomerated with a 460 nm MicroBead difunctionalized latex 55% 5% 100 µeq (4 250 mm) Hydroxide, acetate/ hydroxide 1.0 ml/min 0 90% 3500 psi (24.5 MPa) 0 14 CarboPac PA20 Fast mono-, and disaccharide analysis 6.5 µm diameter nonporous substrate agglomerated with a 130 nm MicroBead quaternary ammonium functionalized latex 55% 5% 65 µeq (3 150 mm) Hydroxide, acetate/ hydroxide 0.5 ml/min 0 100% 3000 psi (21 MPa) 0 14 CarboPac PA100 Oligosaccharide mapping and analysis 8.5 µm diameter nonporous substrate agglomerated with a 275 nm MicroBead di-functionalized latex 55% 6% 90 µeq (4 250 mm) Hydroxide, acetate/ hydroxide 1.0 ml/min 0 90% 4000 psi (28 MPa) 0 14 CarboPac PA200 High resolution oligosaccharide mapping and analysis 5.5 µm diameter nonporous substrate agglomerated with a 43 nm MicroBead quaternary ammonium functionalized latex 55% 6% 35 µeq (3 250 mm) Hydroxide, acetate/ hydroxide 0.5 ml/min 0 100% 4000 psi (28 MPa) 0 14 DNA Column Target Applications Base Matrix Material Substrate Crosslinking Latex Crosslinking Capacity Recommended Eluents Recommended Flow Rate Solvent Compatibility Max. Backpressure ph Range DNAPac PA100 Single stranded DNA or RNA oligonucleotides, restriction fragments, glycoprotein isoforms. 13-µm diameter nonporous substrate agglomerated with a 100-nm MicroBead alkyl quaternary ammonium functionalized latex. 55% 5% 40 µeq Chloride, acetate, bromide, perchlorate: in lithium sodium or ammonium forms 1.5 ml/min 0 100% 4000psi (28MPa) DNAPac PA200 High resolution single stranded DNA or RNA oligonucleotides, restriction fragments, glycoprotein isoforms. 8-µm diameter nonporous substrate agglomerated with a 130-nm MicroBead alkyl quaternary ammonium functionalized latex. 55% 5% 40 µeq Chloride, acetate, bromide, perchlorate: in lithium sodium or ammonium forms 1.2 ml/min 0 100% 4000psi (28MPa) DNASwift 88 Column Selection Guide and Specifications

85 SUPELCOSIL is a trademark of Supelco, Inc. Teflon is a registered trademark of E. I. du Pont de Nemours. Hydromatrix is a trademark of Varian Associates, Inc. Browntec Labs is a trademark and Polypore is a registered trademark of Applied Biosystems. Cyclobond is a registered trademark of Advanced Separation Technologies, Inc. Altima is a trademark of Alltech Associates, Inc. TurbolonSpray is a registered trademark of Applied Biosystems. Soxtec is a trademark of Foss. PEEK is a trademark of Victrex PLC. Milli-Q is a registered trademark of Millipore Corporation. Teramyl is a registered trademark of Novozymes, Denmark. All other trademarks are the property of Thermo Fisher Scientific Inc. and its subsidiaries Thermo Fisher Scientific, Inc. All rights reserved. This information is not intended to encourage use of these products in any manners that might infringe the intellectual property rights of others. Dionex Products: 1228 Titan Way, PO Box 3603, Sunnyvale, CA , (408) North America: U.S./Canada (847) South America: Brazil (55) Europe: Austria (43) , Benelux (31) (32) Denmark (45) , France (33) , Germany (49) Ireland (353) , Italy (39) , Sweden (46) , Switzerland (41) , United Kingdom (44) Asia Pacific: Australia (61) , China (852) , India (91) , Japan (81) , Korea (82) , Singapore (65) , Taiwan (886) LPN /12